to be fair, Mike Chi has never recommended the retrotink for VHS transfer, and in the discord, most people who ask about it are recommended to look into the domesday duplicator instead
Thanks a lot. I watched so many unnecessary "tutorials" that it almost put me off archiving my home videos. In the end I researched about how video signals/capture cards are working and also stayed away from several forums. I was suprised by how you addressed a lot of the same points that I came across myself, especially about TBCs.
You use SRESTORE on every tape? I've been using fpsdivisor=2 or frate=(frame here) depending on what the video looks like. A lot of my tapes have digital elements & ghost frames layered on top of the anime so I usually just go for 29.97 & say (screw it, that looks like the raw mkv did) unless it's very clearly a movie. Having the opening or the credits of an anime break because they shift from 23.976 to 30 just doesn't seem like a good trade off so I switch between the two depending on what it looks like (unless it's very clearly a digital anime, but that's been rare in my experience)Edit 2: Damn, maybe I really should be using deflicker, that might have been bottlenecking me.
Not every tape, but it is handy in certain cases. for the specific Ys LD there is a method using Vapoursynth someone showed me which can deal with really bad ghosting and blending although I'm not sweaty enough at the moment to understand how it works, however, You can specify frame ranges to specific filters, so if you have digital elements at 30fps during frame 100-200, you can basically turn anime IVTC off and process that scene as native 30 to avoid stutters. AVS also supports exporting variable frame rate MKV's and TDecimate also has a VFR mode, which is handy for DVDs. This was the script i used to process the DVD release of Kimagure Orange Road: Summers Beginning, which you might find handy for conventional tape rips aswell if they follow a standard IVTC pattern. LoadPlugin("dgmpgdec2007\x64\DGDecode.dll") DGDecode_D2VSource("[DVD][070523] Shin Kimagure Orange Road ~Soshite, Ano Natsu no Hajimari~\VIDEO_TS\VTS_01_1test.d2v") #################################################### Tfm( mode=2,pp=6,micmatching=3, slow=2, cthresh=8, d2v="[DVD][070523] Shin Kimagure Orange Road ~Soshite, Ano Natsu no Hajimari~\VIDEO_TS\VTS_01_1test.d2v", mchroma=true, mmsco=false) TDecimate(mode=0) ################################################## converttorgb24(interlaced=true).blur(0,1).sharpen(0,1).ConvertToYV24() DotKill_DotKillS(iterations=1) ############################################# a=nnedi3(field=-1, nsize=6, nns=4, qual=2, threads=10) yadifmod2(mode=0, edeint=a) ############################################# Clip=last o =last DH = o.dehalo_alpha(rx=2.5,ry=2.5,darkstr=1.0) bb = DH.removegrain(11) xD = mt_makediff(bb,bb.repair(bb.repair(bb.medianblur(2,-333,-333),1),1)) \ .mt_lut("x 128 - 2.49 * 128 +") xDD = mt_lutxy(xD,mt_makediff(o,DH),"x 128 - y 128 - * 0 < 128 x 128 - abs y 128 - abs < x y ? ?") DH.mt_adddiff(xDD,U=2,V=2) ############################################## crop(4,4,-4,-4) Prefetch(threads=10, frames=10) (FOR VHS YOU CAN PUT THE CROP BEFORE THE DEINTERLACE)
It feels like one of the greater disservices in the video was creating the perception that you need to spend four figures to even get started otherwise. It has the chance of turning away people from even giving it a go otherwise. Clearly, the 4k is top of class when it comes to gaming, but in their attempts to expand their viable market, it has the possibility of doing damage to archival efforts.
My thoughts exactly! and I feel like alot of people are swooned by the idea that if only they had the money to buy the magic box they would have all of their problems solved. But Trying to digitize media is fundamentally a problem created by old hardware limitations so investing heavily in more hardware to solve the hardware problem doesn't make sense to me.
@@The_Smuggler_ Were any of the TBCs you mentioned full frame? I have a TBC-1000 from when they were basically worthless, but those editing decks look really cool. There's like 20 star wipes on some of those things.
Very few TBCs outside of TV station equipment are full frame unfortunately, however a good Field TBC will do quite well. You can improve the performance of a field line TBC inside a video mixer by Buying a VCR or VTR that supports Black Burst Output, Sony SVO series VCRs, amoung others, will output timing information through a B.B. port which most well equipped mixers can interpret and adjust their internal TBC to match the output. This is essentially a genlock signal which acts in tandem with the TBC to further stabilize and Sync the video. it basically lets the VCR talk to the mixer n vice versa. If you use the B.B. signal to record a blank tape it is also possible to perform preprogrammed edits and transitions during A-B editing, since the Mixer knows what the VCR is doing.
I watched all 2 hours and now I have more questions than before. 😅 It was really educational even if I'm a total noob in analog video. I would love to see a video that guides people in the right direction. I started gathering infos and equipments to build a CX card digitizer to be able to pull video out of a VCR's RF signal, but now I have doubts. :D Living in Europe is already a struggle, because I have to learn and find infos about both PAL and NTSC stuff (I use analog NTSC cameras because they are much easier to source and gather, but my family's VHS tapes and everything here are in PAL). Is a "TBC" or a Videomixer necessary for a good capture... Was ever a AV mixer manufactured that was capable of both PAL/SECAM and NTSC? So many questions... Thank you for the video again! I hope you keep making them, because they hold great value!
Thanks for taking the time to watch all the way through! Later this year when I have the time I plan on sitting down n making a follow up to this which covers how I typically process various types of media, including VHS, LDs and DVDs. Until then though I would recommend Reading through the material which is listed on YukiSubsGuide. A lot of it relates to anime but there are many articles linked which explain the fundamentals of NTSC and PAL video. I would say that if the media you are capturing is in good condition a TBC is not necessary for a "Good" capture. TBC's typically help with capture when the film is damaged or degraded in some way. However most home video is damaged in some way, so if you notice wobbling in the frame and dropped frames a TBC can improve the capture. Keep in mind Home video cameras also recorded video in a lower standard compared to a Pre-printed movie tapes so you may not be able to push the upscaling as far as if you were working on something like a laserdisc. Im unsure about multi region video mixers but I believe Roland made PAL Analog mixers at some point and I do know that some Pro-Level VCR's had PAL, SECAM, and NTSC Playback, Panasonic manufactured a few, they were able to do on the fly conversion to NTSC and the Video quality was quite good compared to a consumer level multi region player. Hope that is of some help n thanks for commenting!
@The_Smuggler_ Oh that helped a lot. Thank you, means a lot. You earned a sub after the first 10 minutes and I would definitely watch more videos on this topic by you, as you explain things in a way that is easy to digest. Yeah I understand when is TBC needed I just wasn't sure how much a neccessity it is - can you call a good capture a good capture without TBC, but you explained it here, it's not neccessary it's just needed in a lot of cases. Sony FXE 100 has a PAL counterpart (FXE 100P) and there are a lot of other PAL mixers (Videotechnics, Canera etc.) but something that can handle both PAL and NTSC would be a lifesaver and perhaps money saver. Most of my family's recordings are TV or other captures so the quality should be near broadcast level, not camcorder level. Those are like documentary movies that you cannot find on the internet anymore etc. and my mom would love to watch them again and save them for later. Here in Eastern Europe having an analog videocamera was basically a luxury (Video8, Hi8, or even the Digital8). I was born in 96, but I have zero videos made in my first 12 years as we didn't have a videocamera until then and we are considered to be midclass family by our country's standard. 😅 It's funny how much longer and widespreadedly our country used 35mm photo cameras to capture everything, even in 2005-6, while most of the US and Japan population were already recording stuff on analog or even digital cassettes. Thanks again. Cannot wait to watch your next video! 😇
I have to agree, digitize the closest to source material. And by the way, I don’t think your mouse will live that long, you click so hard. It’s amazing.
I really don't get why people wanted to use the scalers intended for video games for archiving when their intention as a scaler was to be played/watched straight to your display. They're compromise devices in a market where they're limited options if you want a scaler with low input lag. I have a 5x and progressive scan games look great, but for 480i I sometimes just use bob deinterlacing rather than motion adapting since some games it just looks worse than bob. The reason I got it was the auto switcher significantly reduces switching on the display since many older games require that, which is necessary if a display syncs too slow like my JVC projector.
Thank you for the video. It’s a bit long, but there are so many things that could not be covered in five minutes. Everyone will have their opinion, for our part, we prefer a scanned file without corrections with the least compression possible, but it will depend on the expectations of each and the use of the archives once digitized. Not speaking English, please excuse us for a rough translation. The Canal 29 team
I was watching the original video you reviewing here at first because I was interested in the retrotinks. I was also very skeptical about his settings. Good I found your video and I learned a lot. The side by side comparison shows that the I-o usb converter gives really good results. Also I did a lot of research in the last days for finding the best “cheap” way to digitize my vhs collection. At the end it was the ati tv wonder 600, diamond vc 500 and the I-o gv-usb2 which got the best reputations. The I-o is the only one which is easy available. Do you think this is the best for the money or in general which capture device do you prefer? Greeting and keep up the good work
Currently I use the AVerMedia C725B its a very capable card, although it was discontinued a couple years ago. If you can find it on ebay at a decent price I'd recommend it. Although currently for the money the io-data is the best bet, especially if you're just starting out! It has limitations but processing the video correctly with avisynth or hybrid can overcome alot of them and yeild very good results. Id recommend using the money you save to buy a good SVHS, VCR which will do alot in improving the capture quality regardless of the capture card you use.
Hey man, this is a very interesting video! Lots of great info. I'm currently going down the rabbit hole of analog video archival capture (specifically VHS and S-VHS tapes), and I've done more reading than I could ever want to on the different workflows... an old windows XP computer and the period correct hardware and software, or the currently available USB analog capture cards (IO data), or analog to HDMI conversion and HDMI capture card (e.g. retrotink), or analog to SDI conversion and SDI capture, or analog to firewire conversion device then capture (I do have a windows 7 laptop with firewire port), or why not a CRT TV and just play the tapes when I want to watch them :) Anyway, I'm curious as to what your typical process is for capturing analog s-vhs video while retaining the most 'information' from the original source as "reasonably" possible? Not sure I'm smart enough to figure out the radio frequency methods like vhs-decode.
Youll want any capture device that supports native NTSC or PAL video capture, with a 4:2:0 or 4:2:2 color space, Lossless capture. ref NTSC Chroma Sub-Sampling. That Osprey capture card on Amazon seems to be the best non domesday option currently. dont run the video through SDI converters, and don't pay money for a windows Vista TV card. I have tried those methods and I don't know why people have decided those are good options. Anything that changes the format of the video signal before recording should be avoided like the plague. Capture that 4,2,0 video as Uncompressed RGB in Vdub2. then process the video using VaporSynth or Avisynth. Upscale to 4:3 1080p if you're posting on TH-cam, otherwise leave the video at its base resolution. invest in a RAID setup and buy lots of hard drives.
@@The_Smuggler_ I appreciate the reply! Any reason to capture the 486interlaced (visible lines) NTSC video as uncompressed RGB rather than something like lossless compressed YUY2 in 4:2:2? Also, in your opinion, is the IO data GV-USB2 capture device a worthwhile option, or should it be avoided? For what it's worth - I have seen some folks get what I would (maybe naively) consider as pretty damn good results using an analog s-video to SD SDI (525i NTSC, 59.94 fields per second) converter device, then capturing the SDI video in lossless 10-bit 4:2:2, but not sure how it objectively stacks up to the other methods.
Sorry for the confusion I misspoke, By lossless capture I mean using a lossless codec, or in the case of VDUB uncompressed RGB is still capturing 4:2:0 or in the case of broadcast NTSC Video 4:1:1. From my understanding it is impossible to capture interlaced video in 4:4:4, or even in 4:2:2 in many cases since the actual fields are separated into alternating lines and staggered, this halves the chroma, and the Chroma information on NTSC video is halfed again compared to the luma information. 4:2:2 was mostly utilized in DVCAM systems, This was part of the BT601 standard, but the previous standard REC601 was 4:2:0 so If you are capturing movies or anime from the 90s its probably 4:2:0 or in the case of low budget English localizations the dubbing equipment could have brought that down to 4:1:1. While some SDI or converter solutions may let you over sample to 4:2:2 or 4:4:4 in 10 bit, the extra information is just being interpolated. The most important part that many people overlook with SDI converters and Retrotinks, is that the devices are built to assume that the source video is actually properly printed on the tape with standardized Telecine patterns and field timing. If those device misidentify, add or drop a single feild throughout and entire movie, it can screw the IVTC and deinterlace later which is not an issue I've run into using a native ntsc composite capture device.
@@The_Smuggler_ The Osprey capture card offered by "River" Germany unfortunately has no technical specifications other than no audio and 576p, which makes the product 100% useless since PAL is interlaced.
Great video! You should add a message or banner at the start that says: “WARNING: This video contains occasional heavy Canadian-accented words and phrases.”
Interesting take to be sure. I think to properly debunk the actual result, you'd need to have a capture sample by following the tutorial versus what you'd consider to be an ideal method on the same starting content. You'd probably want to zoom way in (like 4x) so that details are more visible on TH-cam without the compression washing out the details on each. The retrotink in this capacity is the actual ADC (analog to digital converter) The HDMI capture card placed after it is really just storing the "already digital" stream it made with various settings depending on how you want to capture. I don't think RetroRGB gets a kickback from Retrotink for these videos, I think they are just trying to show some functionality of a device that many of their user base already has. Yeah, sounds like they could have improved upon the settings, but doesn't mean it isn't a good device for the purpose. While originally designed for gaming, it has an FPGA at its heart which is basically hardware emulation and this is what broadcast TBCs used as well for most of their TBC heavy lifting back in the day. I'd wager modern FPGAs are better on all counts in terms of processing power and efficiency. The "trimming/cropping" that is probably a bit of a misnomer - most likely it is just blanking/masking the "trimmed" areas to solid black. I don't think the retrotink can output a resolution of 1900x1070 as an example, or would do so without further developer setting modifications - most HDMI devices wouldn't like that and may not display that at all, let alone be able to capture it. I assume you have a Sony FXE-120 Mixer there based on how it looks - Checked the data sheet and that uses 8 bit color depth and 4:1:1 chroma subsampling for its processing which negates any advantages of capturing in 10 bit after that. I've seen other similar mixers go as low as 4:1:0 chroma subsampling, so that's something to watch for. Agree that the settings used in the tutorial may not have been ideal, but the functionality could still be there if you were to set things up correctly and if you are a gamer that already has one - so sounds like you could make a better tutorial for the same device haha. The other advantage of having a TBC inside something like the retrotink is fewer analog to digital and digital to analog conversions in the chain. One single analog to digital conversion is ideal. Each one you have is a (hopefully small) "generation loss". Case probably hasn't been made yet that it is the most cost effective or has better performance compared to alternatives as not too many direct comparisons against other popular video chains are out there.
Thanks for the great comment! Later this year I plan to aquire a retrotink and make a more concise and detailed video and try to see how what more detailed comparisons could yield, and what the 480i passthrough mode would look like in terms of an actual avisynth script workflow. He has a follow up video where he uses alot of burnt DVDs and a Wii U to test these cards which I also have issue with. In the case of my current set-up I capture uncompressed YUV, so the file sizes are pretty dumb, since many of my tapes are DVCAM footage which were shot at 4:1:1 in rec 601 which is 8bit I dont get a loss of any color information in theory. the 10 bit Color does however make a big difference when you capture or encode from lossless to H264. Since 10 bit gives the codec more bandwidth for the color you end up with fewer compression related artifacts in the final encode. In my experience encoding 8bit analog as 8bit Digital, almost always results in macro blocks during low light scenes. 10bit color is especially important for Digital Gradients used in late 90s and early 2000s anime where 8bit in H264 attempts to save space by bricking the colors together.
Here is the thing though, Not all people want to keep or capture lossless and process on the computer or have the know how, I would say the majority, Yes the proper way is to seek a good VCR with line TBC and output S-Video, capture into 480i AVI 4:2:2 and keep these master files for future processing. So the video you are criticizing is not technically wrong, it is just one of the ways of doing it, I was criticized for my workflow but I do experiment and learn to find out for myself, this is how it should be, There is no such "The best method to capture analog video tapes", It is goal based not method based.
No, the video I criticized is technically and factually wrong on many points, in particular his complete unfamiliarity with the Telecine Process. including the proper use and basic settings of virtual dub required for synced audio during analog capture. Among other areas. I encourage you to watch further into the video so you can get a better idea of what my criticisms actually are.
@@The_Smuggler_ You do realize that analog video conversion to digital is a two phase process, Capturing and processing. So which part you are criticizing? As far as capturing I don't see that he is wrong about any of it, Now processing which includes but not limited to, de-interlacing, cropping, resizing/upscaling, tone and color correction, noise reduction, encoding. it can be achieved in many different ways, depends on skills and budget, As I mentioned the majority of people are not going to do it your way due to skill limitation.
Over the course of this 2 hour video, I definitely talk about both capturing and processing and I present simple and cheap alternatives, in addition to the AVIsynth processing which I show at the end. If you have any issues with the technical details of what I discussed or if I said anything that is factually incorrect please let me know. But if you just have an issue with how the information was presented then I can't really do anything about that, I'm sorry.
@@The_Smuggler_ I wasn't talking about your video, I was talking about the video you're criticizing, Besides using the retrink for processing I didn't see anything wrong he mentioned in that video.
@@The_Smuggler_I don’t understand the preference for composite over SDI. SVideo into SDI or just Svideo would be a lot better for VHS than composite. Rainbows and and dot crawl really should be handled in hardware and not software.
i transfer lots of VHS tapes using Avermedia ADVC-300 and JVC video corrector CX-7 as well as VHS decode with Dooms day duplicator with RF capture form Video head directly.
I think that there is a happy medium here to be found. I agree that Mr RetroRGB is stepping into a field of expertise that he doesn't have enough knowledge to be an authority on. However, "propaganda" and "shill" are I think not factual statements to make about him. He has affiliate links because that's how you make money in a post adpocalypse TH-cam. He has affiliate links to other products on videos where those products are featured. He also mentions at the start of the video that this solution is meant for people who already own a Tink or to put it another way, don't buy a Tink if your only goal is recording analog media. It's just not the same as, for example, the whole rent a PC scam that Gamer's Nexus recently covered where all these creators were paid directly to give specific talking points about the service, many of which were false. This is a guy who, most likely because he is misinformed or ill-informed, thought that this solution was good and made a video about it. So misleading would be a more apt term IMO (but that's not as clickbaity is it?). The idea that the RetroTink products would be good for movies and tv shows is not something that is prominently promoted from Mike Chi. The only mention of it I could find on the website is this blurb all the way down in the FAQ section; quote "Movies -- the goal of the TINK products has always been games and games first, but we are already adding movie oriented features such as BFI film projection simulation, inverse telecine and 24 Hz restoration." There is no mention here about using it to record either. All that being said, I also think that it isn't crazy in some scenarios to use a RetroTink to digitize analog media. I'll give my own experience as an example. I am a retro gaming enthusiast so I already own a RetroTink for that. It's a godsend for up-scaling retro video games to be played on modern TVs. There are other products that also do a fine job like the OSSC or the venerable Framemeister but I went with the Tink. I also already owned an Elgato capture card because my friends and I thought we would get into streaming years ago and never really did. My family has a decent amount of old home movies and I wanted to archive them so I used the tools I had on hand and was satisfied with the result. Is it the greatest rip of those home movies ever? Probably not, but it's satisfactory for preserving these sentimental moments in a digital format as tapes don't last forever. I think that saying that the output of the RetroTink is "shit" for analog media is something that I disagree with. Is it the best? No. But it's satisfactory. I don't like calling it shit the same way I don't like it when audiophiles shit on people for using MP3s. At the end of the day, if I, the person consuming the media is happy with the results then who cares. I do agree that when you're talking about archiving media, especially media that has never been re-released, then I fully agree that you should use the Doomsday Duplicator and properly decode and format the RAW data after the capture and especially especially when you are going to be sharing that archive with others. But for home, individual use, then it isn't a big deal that it is or isn't the "best." That is perhaps the biggest mistake in the RetroRGB video; not being clear that for true archival purposes you should use something like the doomsday duplicator that captures RAW data that can be decoded and processed later. Finally, I think that being as inflammatory as you were wasn't productive to the conversation around this issue. I get that you are upset but I think that coming to the table in good faith and perhaps even reaching out to RetroRGB or Mike Chi for comment and dialog would have been more productive in moving towards a future where people are more well informed on how to properly archive analog media. In that world, perhaps RetroRGB would release an updated video with better information and/or Mike Chi could make a post to the RetroTink blog making it even more clear that the RetroTink products are not meant for archival backups of analog media. Sorry for the long post but I felt that there were times you were being unfair and wanted to correct those moments. I very much appreciate people like you who have the knowledge and tools to properly archive old/lost media. The work you and people like you do is invaluable.
To be completely clear I agree with much of what you say here. Although to be frank my issues are not which the product itself or Mike Chi, but specifically how the product is being represented in this case. It would be one thing if the tink was shown by itself, however RetroRGB compares it against other workflows using a flawed testing methodology. This actively hurts the commercial viability of these competing products and for a channel which is fairly hardware focused seems like a massive oversight. From where I stand, there is no reason why creators like RetroRGB should be publishing videos, which are this poorly researched. A smaller channel such as my self should not be put in a position where I would need to ask nicely to have revisions made to a video which has already been published and watched by a wide audience. This video is simply about presenting the correct information to the best of my abilities. The inflammatory nature of my speech serves to give this unscripted 2 hour rant some form of rhythm or cadence, and also acts to counter RGB's positive and happy demeanor. You say that this is not indicative of having a good conversation, however we are having this conversation now, and I've had a ton of back and forth in DMs with people asking for clarification, or just thanking me in general for presenting this information. I admit I was unfair and overly salty at many points, but I've also broke down an incredibly complex topic in more or less plain English which people can use as a basis for further study. So take from this video what you will. I made it, It is mine, I have no regrets. PS. I sincerely appreciate you taking the time to comment.
why nobody talk about the HDPVR Rocket ? i use it for analog video capture and it's work great the only fault is that since I am European, I have to go through a device that converts the composite signal into component since the rocket only accepts NTSC composite, in my case I go through a Pioneer DVR-545HX for the conversion
So I bought like the old version of topaz like two years ago they used ffmpeg and just bwdif. their AI block after generally reduced the ssim quality need to write up but yeah maybe the new stuff is better the old stuff sucked tbh.
I really only find the Proteus model produces reliable results on the stuff I've done. Although the results are quite good when it works. The default settings and denoising it does is also way too strong in my opinion, seems to struggle with analog anime especially. For DVDs n digital stuff though it works quite well. If you have Avisynth ive found that some of the resize plug ins can give very good results if you run multiple passes of lighter denoising at certain points in the filter chain.
Topaz is quite slow without cuda acceleration, but i built my computer to brute force stuff so its less of an issue. I'm interested to see where the real time upscaling goes though. Frame Generation and DLSS is working quite well for video games, and faking the frames is more perfomant compared to actually generating them which is a very strange idea to me. Since the Topaz output cooks the video similar to the retrotink method it suffers the same drawback, although to a lesser extent. I can't say I know enough about ffmpeg to compare it to AVS but I will say AVS is quite powerful and at the advanced level can produce very high quality VFR content from DVDs. That said if you are already using Linux and ffmpeg, you should forget AVS and use VaporSynth instead. it's much more mordern and is under active development. the only reason I've stayed on AVS this long is because the vast amount of old forum posts which relate to DVD and VHS encoding has been incredibly helpful as a learning tool for myself.
Gamers are anal! Retro RGB is aimed at games and Gamers know they need HDMI capture!! The Retro Tink does what it does really well, it's a device which does real time analogue signal clocking and de-interlacing with a plethora of options! It does it all in real time with as close to 0 lag as possible an gets the accuracy games crave and outputs it in HDMI for modern TV's. The RetroTink has so many technical options and is so versatile it's ridiculous, you can fuck things up as much as you can fix it. the Real Question is what are you are to trying to do? just getting something that looks good in real time on a TV, or capture the best raw footage in it's original format for archiving manipulating on a computer. the Retro Tink does both and often does it way better than most of those insanely priced video processors.
![Путин и ПЕСКОВ ждут звонка от ТРАМПА 😁 [Пародия]](/img/n.gif)
to be fair, Mike Chi has never recommended the retrotink for VHS transfer, and in the discord, most people who ask about it are recommended to look into the domesday duplicator instead
That sort of advice is going from bad to worse.
Thanks a lot. I watched so many unnecessary "tutorials" that it almost put me off archiving my home videos. In the end I researched about how video signals/capture cards are working and also stayed away from several forums. I was suprised by how you addressed a lot of the same points that I came across myself, especially about TBCs.
You use SRESTORE on every tape? I've been using fpsdivisor=2 or frate=(frame here) depending on what the video looks like. A lot of my tapes have digital elements & ghost frames layered on top of the anime so I usually just go for 29.97 & say (screw it, that looks like the raw mkv did) unless it's very clearly a movie. Having the opening or the credits of an anime break because they shift from 23.976 to 30 just doesn't seem like a good trade off so I switch between the two depending on what it looks like (unless it's very clearly a digital anime, but that's been rare in my experience)Edit 2: Damn, maybe I really should be using deflicker, that might have been bottlenecking me.
Not every tape, but it is handy in certain cases. for the specific Ys LD there is a method using Vapoursynth someone showed me which can deal with really bad ghosting and blending although I'm not sweaty enough at the moment to understand how it works, however,
You can specify frame ranges to specific filters, so if you have digital elements at 30fps during frame 100-200, you can basically turn anime IVTC off and process that scene as native 30 to avoid stutters. AVS also supports exporting variable frame rate MKV's and TDecimate also has a VFR mode, which is handy for DVDs.
This was the script i used to process the DVD release of Kimagure Orange Road: Summers Beginning, which you might find handy for conventional tape rips aswell if they follow a standard IVTC pattern.
LoadPlugin("dgmpgdec2007\x64\DGDecode.dll")
DGDecode_D2VSource("[DVD][070523] Shin Kimagure Orange Road ~Soshite, Ano Natsu no Hajimari~\VIDEO_TS\VTS_01_1test.d2v")
####################################################
Tfm( mode=2,pp=6,micmatching=3, slow=2, cthresh=8, d2v="[DVD][070523] Shin Kimagure Orange Road ~Soshite, Ano Natsu no Hajimari~\VIDEO_TS\VTS_01_1test.d2v", mchroma=true, mmsco=false)
TDecimate(mode=0)
##################################################
converttorgb24(interlaced=true).blur(0,1).sharpen(0,1).ConvertToYV24()
DotKill_DotKillS(iterations=1)
#############################################
a=nnedi3(field=-1, nsize=6, nns=4, qual=2, threads=10)
yadifmod2(mode=0, edeint=a)
#############################################
Clip=last
o =last
DH = o.dehalo_alpha(rx=2.5,ry=2.5,darkstr=1.0)
bb = DH.removegrain(11)
xD = mt_makediff(bb,bb.repair(bb.repair(bb.medianblur(2,-333,-333),1),1))
\ .mt_lut("x 128 - 2.49 * 128 +")
xDD = mt_lutxy(xD,mt_makediff(o,DH),"x 128 - y 128 - * 0 < 128 x 128 - abs y 128 - abs < x y ? ?")
DH.mt_adddiff(xDD,U=2,V=2)
##############################################
crop(4,4,-4,-4)
Prefetch(threads=10, frames=10)
(FOR VHS YOU CAN PUT THE CROP BEFORE THE DEINTERLACE)
It feels like one of the greater disservices in the video was creating the perception that you need to spend four figures to even get started otherwise. It has the chance of turning away people from even giving it a go otherwise.
Clearly, the 4k is top of class when it comes to gaming, but in their attempts to expand their viable market, it has the possibility of doing damage to archival efforts.
My thoughts exactly! and I feel like alot of people are swooned by the idea that if only they had the money to buy the magic box they would have all of their problems solved. But Trying to digitize media is fundamentally a problem created by old hardware limitations so investing heavily in more hardware to solve the hardware problem doesn't make sense to me.
@@The_Smuggler_ Were any of the TBCs you mentioned full frame? I have a TBC-1000 from when they were basically worthless, but those editing decks look really cool. There's like 20 star wipes on some of those things.
Very few TBCs outside of TV station equipment are full frame unfortunately, however a good Field TBC will do quite well.
You can improve the performance of a field line TBC inside a video mixer by Buying a VCR or VTR that supports Black Burst Output,
Sony SVO series VCRs, amoung others, will output timing information through a B.B. port which most well equipped mixers can interpret and adjust their internal TBC to match the output. This is essentially a genlock signal which acts in tandem with the TBC to further stabilize and Sync the video. it basically lets the VCR talk to the mixer n vice versa.
If you use the B.B. signal to record a blank tape it is also possible to perform preprogrammed edits and transitions during A-B editing, since the Mixer knows what the VCR is doing.
I watched all 2 hours and now I have more questions than before. 😅 It was really educational even if I'm a total noob in analog video. I would love to see a video that guides people in the right direction. I started gathering infos and equipments to build a CX card digitizer to be able to pull video out of a VCR's RF signal, but now I have doubts. :D Living in Europe is already a struggle, because I have to learn and find infos about both PAL and NTSC stuff (I use analog NTSC cameras because they are much easier to source and gather, but my family's VHS tapes and everything here are in PAL). Is a "TBC" or a Videomixer necessary for a good capture... Was ever a AV mixer manufactured that was capable of both PAL/SECAM and NTSC? So many questions...
Thank you for the video again! I hope you keep making them, because they hold great value!
Thanks for taking the time to watch all the way through! Later this year when I have the time I plan on sitting down n making a follow up to this which covers how I typically process various types of media, including VHS, LDs and DVDs. Until then though I would recommend Reading through the material which is listed on YukiSubsGuide. A lot of it relates to anime but there are many articles linked which explain the fundamentals of NTSC and PAL video.
I would say that if the media you are capturing is in good condition a TBC is not necessary for a "Good" capture. TBC's typically help with capture when the film is damaged or degraded in some way. However most home video is damaged in some way, so if you notice wobbling in the frame and dropped frames a TBC can improve the capture.
Keep in mind Home video cameras also recorded video in a lower standard compared to a Pre-printed movie tapes so you may not be able to push the upscaling as far as if you were working on something like a laserdisc.
Im unsure about multi region video mixers but I believe Roland made PAL Analog mixers at some point and I do know that some Pro-Level VCR's had PAL, SECAM, and NTSC Playback, Panasonic manufactured a few, they were able to do on the fly conversion to NTSC and the Video quality was quite good compared to a consumer level multi region player.
Hope that is of some help n thanks for commenting!
@The_Smuggler_ Oh that helped a lot. Thank you, means a lot. You earned a sub after the first 10 minutes and I would definitely watch more videos on this topic by you, as you explain things in a way that is easy to digest.
Yeah I understand when is TBC needed I just wasn't sure how much a neccessity it is - can you call a good capture a good capture without TBC, but you explained it here, it's not neccessary it's just needed in a lot of cases.
Sony FXE 100 has a PAL counterpart (FXE 100P) and there are a lot of other PAL mixers (Videotechnics, Canera etc.) but something that can handle both PAL and NTSC would be a lifesaver and perhaps money saver.
Most of my family's recordings are TV or other captures so the quality should be near broadcast level, not camcorder level. Those are like documentary movies that you cannot find on the internet anymore etc. and my mom would love to watch them again and save them for later. Here in Eastern Europe having an analog videocamera was basically a luxury (Video8, Hi8, or even the Digital8). I was born in 96, but I have zero videos made in my first 12 years as we didn't have a videocamera until then and we are considered to be midclass family by our country's standard. 😅 It's funny how much longer and widespreadedly our country used 35mm photo cameras to capture everything, even in 2005-6, while most of the US and Japan population were already recording stuff on analog or even digital cassettes.
Thanks again. Cannot wait to watch your next video! 😇
I have to agree, digitize the closest to source material. And by the way, I don’t think your mouse will live that long, you click so hard. It’s amazing.
I really don't get why people wanted to use the scalers intended for video games for archiving when their intention as a scaler was to be played/watched straight to your display. They're compromise devices in a market where they're limited options if you want a scaler with low input lag. I have a 5x and progressive scan games look great, but for 480i I sometimes just use bob deinterlacing rather than motion adapting since some games it just looks worse than bob. The reason I got it was the auto switcher significantly reduces switching on the display since many older games require that, which is necessary if a display syncs too slow like my JVC projector.
Thank you for the video. It’s a bit long, but there are so many things that could not be covered in five minutes. Everyone will have their opinion, for our part, we prefer a scanned file without corrections with the least compression possible, but it will depend on the expectations of each and the use of the archives once digitized. Not speaking English, please excuse us for a rough translation.
The Canal 29 team
Im happy i could provide helpful information! Thank you for taking the time to comment i appreciate it!
I was watching the original video you reviewing here at first because I was interested in the retrotinks. I was also very skeptical about his settings. Good I found your video and I learned a lot. The side by side comparison shows that the I-o usb converter gives really good results. Also I did a lot of research in the last days for finding the best “cheap” way to digitize my vhs collection. At the end it was the ati tv wonder 600, diamond vc 500 and the I-o gv-usb2 which got the best reputations. The I-o is the only one which is easy available. Do you think this is the best for the money or in general which capture device do you prefer? Greeting and keep up the good work
Currently I use the AVerMedia C725B its a very capable card, although it was discontinued a couple years ago. If you can find it on ebay at a decent price I'd recommend it.
Although currently for the money the io-data is the best bet, especially if you're just starting out!
It has limitations but processing the video correctly with avisynth or hybrid can overcome alot of them and yeild very good results.
Id recommend using the money you save to buy a good SVHS, VCR which will do alot in improving the capture quality regardless of the capture card you use.
Hey man, this is a very interesting video! Lots of great info. I'm currently going down the rabbit hole of analog video archival capture (specifically VHS and S-VHS tapes), and I've done more reading than I could ever want to on the different workflows... an old windows XP computer and the period correct hardware and software, or the currently available USB analog capture cards (IO data), or analog to HDMI conversion and HDMI capture card (e.g. retrotink), or analog to SDI conversion and SDI capture, or analog to firewire conversion device then capture (I do have a windows 7 laptop with firewire port), or why not a CRT TV and just play the tapes when I want to watch them :) Anyway, I'm curious as to what your typical process is for capturing analog s-vhs video while retaining the most 'information' from the original source as "reasonably" possible? Not sure I'm smart enough to figure out the radio frequency methods like vhs-decode.
Youll want any capture device that supports native NTSC or PAL video capture, with a 4:2:0 or 4:2:2 color space, Lossless capture. ref NTSC Chroma Sub-Sampling.
That Osprey capture card on Amazon seems to be the best non domesday option currently.
dont run the video through SDI converters, and don't pay money for a windows Vista TV card. I have tried those methods and I don't know why people have decided those are good options. Anything that changes the format of the video signal before recording should be avoided like the plague.
Capture that 4,2,0 video as Uncompressed RGB in Vdub2.
then process the video using VaporSynth or Avisynth.
Upscale to 4:3 1080p if you're posting on TH-cam, otherwise leave the video at its base resolution.
invest in a RAID setup and buy lots of hard drives.
@@The_Smuggler_ I appreciate the reply! Any reason to capture the 486interlaced (visible lines) NTSC video as uncompressed RGB rather than something like lossless compressed YUY2 in 4:2:2? Also, in your opinion, is the IO data GV-USB2 capture device a worthwhile option, or should it be avoided?
For what it's worth - I have seen some folks get what I would (maybe naively) consider as pretty damn good results using an analog s-video to SD SDI (525i NTSC, 59.94 fields per second) converter device, then capturing the SDI video in lossless 10-bit 4:2:2, but not sure how it objectively stacks up to the other methods.
Sorry for the confusion I misspoke, By lossless capture I mean using a lossless codec, or in the case of VDUB uncompressed RGB is still capturing 4:2:0 or in the case of broadcast NTSC Video 4:1:1.
From my understanding it is impossible to capture interlaced video in 4:4:4, or even in 4:2:2 in many cases since the actual fields are separated into alternating lines and staggered, this halves the chroma, and the Chroma information on NTSC video is halfed again compared to the luma information.
4:2:2 was mostly utilized in DVCAM systems, This was part of the BT601 standard, but the previous standard REC601 was 4:2:0 so If you are capturing movies or anime from the 90s its probably 4:2:0 or in the case of low budget English localizations the dubbing equipment could have brought that down to 4:1:1.
While some SDI or converter solutions may let you over sample to 4:2:2 or 4:4:4 in 10 bit, the extra information is just being interpolated.
The most important part that many people overlook with SDI converters and Retrotinks, is that the devices are built to assume that the source video is actually properly printed on the tape with standardized Telecine patterns and field timing. If those device misidentify, add or drop a single feild throughout and entire movie, it can screw the IVTC and deinterlace later which is not an issue I've run into using a native ntsc composite capture device.
@@The_Smuggler_ The Osprey capture card offered by "River" Germany unfortunately has no technical specifications other than no audio and 576p, which makes the product 100% useless since PAL is interlaced.
Great video! You should add a message or banner at the start that says: “WARNING: This video contains occasional heavy Canadian-accented words and phrases.”
Interesting take to be sure.
I think to properly debunk the actual result, you'd need to have a capture sample by following the tutorial versus what you'd consider to be an ideal method on the same starting content. You'd probably want to zoom way in (like 4x) so that details are more visible on TH-cam without the compression washing out the details on each.
The retrotink in this capacity is the actual ADC (analog to digital converter) The HDMI capture card placed after it is really just storing the "already digital" stream it made with various settings depending on how you want to capture.
I don't think RetroRGB gets a kickback from Retrotink for these videos, I think they are just trying to show some functionality of a device that many of their user base already has. Yeah, sounds like they could have improved upon the settings, but doesn't mean it isn't a good device for the purpose. While originally designed for gaming, it has an FPGA at its heart which is basically hardware emulation and this is what broadcast TBCs used as well for most of their TBC heavy lifting back in the day. I'd wager modern FPGAs are better on all counts in terms of processing power and efficiency.
The "trimming/cropping" that is probably a bit of a misnomer - most likely it is just blanking/masking the "trimmed" areas to solid black. I don't think the retrotink can output a resolution of 1900x1070 as an example, or would do so without further developer setting modifications - most HDMI devices wouldn't like that and may not display that at all, let alone be able to capture it.
I assume you have a Sony FXE-120 Mixer there based on how it looks - Checked the data sheet and that uses 8 bit color depth and 4:1:1 chroma subsampling for its processing which negates any advantages of capturing in 10 bit after that. I've seen other similar mixers go as low as 4:1:0 chroma subsampling, so that's something to watch for.
Agree that the settings used in the tutorial may not have been ideal, but the functionality could still be there if you were to set things up correctly and if you are a gamer that already has one - so sounds like you could make a better tutorial for the same device haha. The other advantage of having a TBC inside something like the retrotink is fewer analog to digital and digital to analog conversions in the chain. One single analog to digital conversion is ideal. Each one you have is a (hopefully small) "generation loss". Case probably hasn't been made yet that it is the most cost effective or has better performance compared to alternatives as not too many direct comparisons against other popular video chains are out there.
Thanks for the great comment!
Later this year I plan to aquire a retrotink and make a more concise and detailed video and try to see how what more detailed comparisons could yield, and what the 480i passthrough mode would look like in terms of an actual avisynth script workflow.
He has a follow up video where he uses alot of burnt DVDs and a Wii U to test these cards which I also have issue with.
In the case of my current set-up I capture uncompressed YUV, so the file sizes are pretty dumb, since many of my tapes are DVCAM footage which were shot at 4:1:1 in rec 601 which is 8bit I dont get a loss of any color information in theory.
the 10 bit Color does however make a big difference when you capture or encode from lossless to H264.
Since 10 bit gives the codec more bandwidth for the color you end up with fewer compression related artifacts in the final encode. In my experience encoding 8bit analog as 8bit Digital, almost always results in macro blocks during low light scenes. 10bit color is especially important for Digital Gradients used in late 90s and early 2000s anime where 8bit in H264 attempts to save space by bricking the colors together.
I wish I saw this video before I bought the tink 4k 🤦♂️ All that money down the drain. Thanks for uploading this vid.
Here is the thing though, Not all people want to keep or capture lossless and process on the computer or have the know how, I would say the majority, Yes the proper way is to seek a good VCR with line TBC and output S-Video, capture into 480i AVI 4:2:2 and keep these master files for future processing. So the video you are criticizing is not technically wrong, it is just one of the ways of doing it, I was criticized for my workflow but I do experiment and learn to find out for myself, this is how it should be, There is no such "The best method to capture analog video tapes", It is goal based not method based.
No, the video I criticized is technically and factually wrong on many points, in particular his complete unfamiliarity with the Telecine Process. including the proper use and basic settings of virtual dub required for synced audio during analog capture. Among other areas. I encourage you to watch further into the video so you can get a better idea of what my criticisms actually are.
@@The_Smuggler_ You do realize that analog video conversion to digital is a two phase process, Capturing and processing. So which part you are criticizing? As far as capturing I don't see that he is wrong about any of it, Now processing which includes but not limited to, de-interlacing, cropping, resizing/upscaling, tone and color correction, noise reduction, encoding. it can be achieved in many different ways, depends on skills and budget, As I mentioned the majority of people are not going to do it your way due to skill limitation.
Over the course of this 2 hour video, I definitely talk about both capturing and processing and I present simple and cheap alternatives, in addition to the AVIsynth processing which I show at the end.
If you have any issues with the technical details of what I discussed or if I said anything that is factually incorrect please let me know. But if you just have an issue with how the information was presented then I can't really do anything about that, I'm sorry.
@@The_Smuggler_ I wasn't talking about your video, I was talking about the video you're criticizing, Besides using the retrink for processing I didn't see anything wrong he mentioned in that video.
@@The_Smuggler_I don’t understand the preference for composite over SDI. SVideo into SDI or just Svideo would be a lot better for VHS than composite.
Rainbows and and dot crawl really should be handled in hardware and not software.
Smug going hard in the paint on this one!
I am just glad my few 90s anime I have on VHS got all great blu ray releases. Akira, venus wars and crying freeman 😁
i transfer lots of VHS tapes using Avermedia ADVC-300 and JVC video corrector CX-7 as well as VHS decode with Dooms day duplicator with RF capture form Video head directly.
I think that there is a happy medium here to be found. I agree that Mr RetroRGB is stepping into a field of expertise that he doesn't have enough knowledge to be an authority on. However, "propaganda" and "shill" are I think not factual statements to make about him. He has affiliate links because that's how you make money in a post adpocalypse TH-cam. He has affiliate links to other products on videos where those products are featured. He also mentions at the start of the video that this solution is meant for people who already own a Tink or to put it another way, don't buy a Tink if your only goal is recording analog media. It's just not the same as, for example, the whole rent a PC scam that Gamer's Nexus recently covered where all these creators were paid directly to give specific talking points about the service, many of which were false. This is a guy who, most likely because he is misinformed or ill-informed, thought that this solution was good and made a video about it. So misleading would be a more apt term IMO (but that's not as clickbaity is it?). The idea that the RetroTink products would be good for movies and tv shows is not something that is prominently promoted from Mike Chi. The only mention of it I could find on the website is this blurb all the way down in the FAQ section; quote "Movies -- the goal of the TINK products has always been games and games first, but we are already adding movie oriented features such as BFI film projection simulation, inverse telecine and 24 Hz restoration." There is no mention here about using it to record either. All that being said, I also think that it isn't crazy in some scenarios to use a RetroTink to digitize analog media. I'll give my own experience as an example. I am a retro gaming enthusiast so I already own a RetroTink for that. It's a godsend for up-scaling retro video games to be played on modern TVs. There are other products that also do a fine job like the OSSC or the venerable Framemeister but I went with the Tink. I also already owned an Elgato capture card because my friends and I thought we would get into streaming years ago and never really did. My family has a decent amount of old home movies and I wanted to archive them so I used the tools I had on hand and was satisfied with the result. Is it the greatest rip of those home movies ever? Probably not, but it's satisfactory for preserving these sentimental moments in a digital format as tapes don't last forever. I think that saying that the output of the RetroTink is "shit" for analog media is something that I disagree with. Is it the best? No. But it's satisfactory. I don't like calling it shit the same way I don't like it when audiophiles shit on people for using MP3s. At the end of the day, if I, the person consuming the media is happy with the results then who cares. I do agree that when you're talking about archiving media, especially media that has never been re-released, then I fully agree that you should use the Doomsday Duplicator and properly decode and format the RAW data after the capture and especially especially when you are going to be sharing that archive with others. But for home, individual use, then it isn't a big deal that it is or isn't the "best." That is perhaps the biggest mistake in the RetroRGB video; not being clear that for true archival purposes you should use something like the doomsday duplicator that captures RAW data that can be decoded and processed later. Finally, I think that being as inflammatory as you were wasn't productive to the conversation around this issue. I get that you are upset but I think that coming to the table in good faith and perhaps even reaching out to RetroRGB or Mike Chi for comment and dialog would have been more productive in moving towards a future where people are more well informed on how to properly archive analog media. In that world, perhaps RetroRGB would release an updated video with better information and/or Mike Chi could make a post to the RetroTink blog making it even more clear that the RetroTink products are not meant for archival backups of analog media. Sorry for the long post but I felt that there were times you were being unfair and wanted to correct those moments. I very much appreciate people like you who have the knowledge and tools to properly archive old/lost media. The work you and people like you do is invaluable.
Well said
To be completely clear I agree with much of what you say here.
Although to be frank my issues are not which the product itself or Mike Chi, but specifically how the product is being represented in this case. It would be one thing if the tink was shown by itself, however RetroRGB compares it against other workflows using a flawed testing methodology. This actively hurts the commercial viability of these competing products and for a channel which is fairly hardware focused seems like a massive oversight.
From where I stand, there is no reason why creators like RetroRGB should be publishing videos, which are this poorly researched. A smaller channel such as my self should not be put in a position where I would need to ask nicely to have revisions made to a video which has already been published and watched by a wide audience.
This video is simply about presenting the correct information to the best of my abilities. The inflammatory nature of my speech serves to give this unscripted 2 hour rant some form of rhythm or cadence, and also acts to counter RGB's positive and happy demeanor. You say that this is not indicative of having a good conversation, however we are having this conversation now, and I've had a ton of back and forth in DMs with people asking for clarification, or just thanking me in general for presenting this information.
I admit I was unfair and overly salty at many points, but I've also broke down an incredibly complex topic in more or less plain English which people can use as a basis for further study. So take from this video what you will. I made it, It is mine, I have no regrets.
PS. I sincerely appreciate you taking the time to comment.
why nobody talk about the HDPVR Rocket ?
i use it for analog video capture and it's work great
the only fault is that since I am European, I have to go through a device that converts the composite signal into component since the rocket only accepts NTSC composite, in my case I go through a Pioneer DVR-545HX for the conversion
I'd rather try to get the 4:4:4 HDMI output from the Pioneer DVR-560H to work with Blackmagic UltraStudio Recorder 3G.
So I bought like the old version of topaz like two years ago they used ffmpeg and just bwdif. their AI block after generally reduced the ssim quality need to write up but yeah maybe the new stuff is better the old stuff sucked tbh.
I really only find the Proteus model produces reliable results on the stuff I've done. Although the results are quite good when it works.
The default settings and denoising it does is also way too strong in my opinion, seems to struggle with analog anime especially.
For DVDs n digital stuff though it works quite well. If you have Avisynth ive found that some of the resize plug ins can give very good results if you run multiple passes of lighter denoising at certain points in the filter chain.
@ I always found the avisynth setup a pain compared to ffmpeg but I mainly use Linux. Topaz takes forever
Topaz is quite slow without cuda acceleration, but i built my computer to brute force stuff so its less of an issue. I'm interested to see where the real time upscaling goes though.
Frame Generation and DLSS is working quite well for video games, and faking the frames is more perfomant compared to actually generating them which is a very strange idea to me.
Since the Topaz output cooks the video similar to the retrotink method it suffers the same drawback, although to a lesser extent.
I can't say I know enough about ffmpeg to compare it to AVS but I will say AVS is quite powerful and at the advanced level can produce very high quality VFR content from DVDs.
That said if you are already using Linux and ffmpeg, you should forget AVS and use VaporSynth instead. it's much more mordern and is under active development. the only reason I've stayed on AVS this long is because the vast amount of old forum posts which relate to DVD and VHS encoding has been incredibly helpful as a learning tool for myself.
I wish I had recorded something, so I could digitize something now. 😢
YESSSSS, I'VE BEEN THINKING ABOUT DOING THIS. Thank you for covering the issue with the retrotink propaganda.
Gamers are anal! Retro RGB is aimed at games and Gamers know they need HDMI capture!! The Retro Tink does what it does really well, it's a device which does real time analogue signal clocking and de-interlacing with a plethora of options! It does it all in real time with as close to 0 lag as possible an gets the accuracy games crave and outputs it in HDMI for modern TV's.
The RetroTink has so many technical options and is so versatile it's ridiculous, you can fuck things up as much as you can fix it. the Real Question is what are you are to trying to do? just getting something that looks good in real time on a TV, or capture the best raw footage in it's original format for archiving manipulating on a computer. the Retro Tink does both and often does it way better than most of those insanely priced video processors.
this is what news channels should be airing
Wrong