I found a mistake in the spreadsheet and corrected it. Specifically, the dual card numbers were off on the UHS II + UHS I. It takes 61 seconds to clear vs 34 seconds. Sorry about that. I've updated the spreadsheet download to reflect the new numbers.
Hi Rob when you show the speed in which it achieves the tests covered the camera is amazing. The transfer of data from the sensor to a ready image is incredible don't you think. For me also the camera controlling focusing sensor shift to stabilise and everthing else makes the tool a work of art in itself, if I try to comprehend it. Your effort in the work you do along with the good it provides as always is just great Rob. Thank you.
Hi Rob. I watched several of your videos yesterday, including this one. I think it was here that you mentioned that in some cases, very fast sequential shots are taken too rapidly for image stabilisation to work. If so, I just noticed in the pdf manual for my M1 ii this: pdf page 114: cog menu: custom menu C2: Image Stabilization "Sets the priority function during sequential shooting. [Fps Priority]: Shooting speed gets priority over image stabilization. The sensor will not be reset to the center during sequential shooting. [IS Priority]: Image stabilization gets priority over shooting speed. The sensor will be reset to the center per frame of sequential shooting. The shooting speed will drop slightly." Cheers for all your great advice - my confidence is gaining momentum!
Wow! You sure took a lot of time and effort to examine this! It is very useful for those who may already have SD cards and do not need the extra performance. Another benefit of M43! Thanks again for sharing!
I appreciate you also showing what settings you typically use, inside the speed comparisons and settings juxtapositions. And thanks for all the technical explanations.
Great job indeed! I am currently using an em5 miii and i shoot a lot of photos of animals in zoos. I never use very high burst modes. Because, except if your are a fan of humming birds, 10 fps are enough in main cases. In addition, using burst mode lets you go home after three hours in a zoo with something like 1000 photos. I can t imagine if i was using 120fps. But i must admit that in some circumstances (capture pro) it may be useful. Anyway, i live in France (it explains my bad english) and OM-1 will be avaliable only next week, i'll buy it as soon as it will be in my favorite store. Thanks again for your videos.
This would have been just as useful as your excellent standard, had I the funds to purchase this camera. In line with your usual practical and very high standard. Enjoyed watching, even not having the camera!
As I understand it, the buffer is actually emptying as fast as it can to the card and is dependent upon the speed the card can be written to, which varies from card brand/model to card. So it's not like the buffer fills up and then over flows into the memory card. It's constantly emptying to the card(s) once the shutter is fully pressed. So if a particular file format and card spec can be written to faster or as fast as the camera is recording, then conceivably, we could just shoot at that format and speed until the card's memory was full. With some cards, if recording at the extreme fps speeds beyond 20 or 30 fps, then, depending on how fast the card can be written to, the results may vary somewhat. But I found with most of our cameras, that write speed was more limited by the camera hardware than the particular card, beyond a certain threshold. For 12MB jpegs, it's possible to have a card that can be written to at 20fps and never fill the buffer, if the card can be written to fast enough. But cards are not all equal with respect to the kind of SD card hardware and firmware a particular camera employs. I remember that the OM-D E-M1 Mark II could not write faster than 180MB/s no matter what card we used. I'm not sure what the speed for the OM-1 might be. The OM-D E-M1 Mark III was about twice as fast as the OM-D E-M1 Mark II in my tests.
@@RobTrek there are a lot of variables but you did a really good job. I have done my own similar tests and your numbers look almost identical to my results. I sort of imagine that the faster the action I want to record, and the more fps I need, the shorter the duration I have to cover. The slower the action, the longer the duration but the less fps I need so it seems to work out fine and I can’t remember many times in recent years where I was hitting the buffer limits before I was done shooting.
Thank you for doing this. I have spent money on Sony Tough UHS-II cards and didn't notice any real change between them and my PNY or Angelbird UHS-I cards. But a HUGE difference is that I can shoot for longer with 256GB & 512GB UHS-I cards than I can shoot with the 64GB Sonys that I have to swap-out multiple times...(And still spend less money!) Also, is it just me, or did you actually find that the UHS-I cards offloaded FASTER than the UHS-II cards? In your timer sequences you show that the UHS-I card took 3:24.68 while the UHS-II took 4:25.19. Seems like it should be the other way-around? Faster cards SHOULD offload faster if a true UHS-II reader is used and the SSD being written to can keep up... Right? I could see if the reader only does UHS-I, and then the speed gains of UHS-II cards would be negated and even might be a little slower. Just asking, and NOT a criticism.
Ugh! I got the timings reversed. Sorry about that. I did use UHS-II reader to an NVME SSD. And yes, sometimes a UHS-II card will read slower in UHS-I card readers. My older Sandisk UHS-II card reads at about 1/2 the speed of a good UHS-I card in a UHS-1 reader. However, my newer ProGrade UHS-II cards read at the same speed as UHS-I cards in a UHS-1 reader.
Wow! What a ton of work you did. I’ve been searching the internet for similar information. Thanks! QUESTION; did you notice any difference in V30 vs. V90? I’m guessing that you didn’t based on the photo of your memory cards. At least now you have cards for your second OM-1:)
I was going to ask the same question. I have a couple V90 and a couple v60 cards, but is it worth swiping them out of my Sony FF body? Thanks for doing all the work Rob. Very helpful
When using a single card slot, shooting Raw+Jpg, it made a big difference. From there, the differences were less noticeable, especially when using dual card slots.
You not only try, but always succeed. Olympus and OMS owners owe you a great debt. Have still yet to really master my Pen F, however, your excellent Pen F videos have been a great aid- manual is rather useless. Btw, whilst rarely do any post, have the Pen F set to Raw, Superfine ( thanks to your video) and jpeg. The day may come when I will get someone to do some post tweaking on an image, hence the Raw setting for each image. The downside is the time taken to read to Sd cards. I only use Scandisk. Am aware that there are now very fast ( and expensive) cards ,maybe more for the mega sized files for the new mirrorless huge files. Am willing to purchase faster reading cards, but there may be an inbuilt limit as to how quickly the camera can read to any given card, no matter how fast it's read speed . Any tips, or links to a video would be greatly appreciated.
Thanks for testing out so many scenarios. Many reported the buffer clearing time is twice longer at ISO higher than 1600. My quick check with RAW only and single UHS-II seems to reflect that. This could explain the strange results in your spreadsheet where you had 71s for single UHS-I card but 33s for dual UHS-I at 20fps RAW+J. My quick take is: Shooting RAW only is the only way to take advantage of UHS-II. Shooting RAW+J increases processing time so much that UHS-I is good enough.
That's interesting. All my tests were done at iso200, except one set at 1600. However, the SQ20 set was done at iso200. I just tested UHSI R+J SQ20 again and got the same result at iso200 (103 images at 71sec to clear), but did it again at iso 3200 and it took 103 images and 83 seconds to clear. So it seems iso processing is putting a drag on the write times. Thanks. Also, you've found a mistake I made in the spreadsheet and I've updated it. It takes 61 seconds to clear when using dual card slots with UHS-I in one or both slots.
Hi Rob, very elaborate and certainly useful. I don't think that shooting for 10sec is very realistic, at least for adventure, sports or wildlife, which I shoot regularly. Horse in races, run avg. 50Km/h (~30mph), so in 1 sec horses have moved ~14m (46ft) and a burst of 1 to 2 sec, possibly panning, is actually long. Using ProCap with E-M1 III and now OM-1, I find that 20-50 frames are more than adequate to capture the moment. Need less frames if using super telephoto like 300/4 Pro to keep subject in the frame. Need more frames using wider view to capture butterfly taking off with 40-150/2.8 Pro. Capturing more frames is probably adding frames of little, if any, use. Therefore, IMHO the buffers of E-M1 III, E-M1X and OM-1 are in practice more than adequate. With this thinking, UHS-II cards with the highest write about 250-280Mbps are very useful to empty the buffer quickly, These cameras actually allow to capture a new sequence before the buffer is empty, so speed is even more useful.
How did you measure the buffer speed. The buffer is an internal camera memory and the read/write speed should be always the same. When the buffer is full then latest the writing on SD card starts. And writing speed of SD cards differs a lot by specification, though also influenced by camera interface.
The buffer seems to capture 93fps. I took a picture of a timer, then counted the number of frames each second. For example, at 120fps, I counted 93 frames at the 1 second mark, then 6 frames in the 2 second mark and so on. There is some margin of error but I did the tests multiple times and was consistently within a few frames. Near the end of the the video, I showed the cards I used and a standard crystal diskmark score.
@@RobTrek I find it varies a little bit depending on the card since the buffer is emptying continuously once the shutter is fully pressed so if a particular card can be written to faster than another card, the results will reflect that.
Thank you ! Very useful information. I have two cards: uhs I and uhs II, If I only shoot raw and use one card at each time, so my best choice is to save to the uhs II card first, am I right? Thank you very much!
Fast cards are worth the money if you are a burst shooter. The upload speed is just a bonus. Thanks for the data validation. I didn't need to be sold, but am certainly vindicated in springing for the extra cost of fast cards.
Thank you for bringing us this information. What struck me is that UHS-I and UHS-II are almost identical. I tried to mimic your test wth my SD cards and I get about the same performance. What I didn't figure out is how you find the FPS value with buffer full. I used to record the shutter sound but that isn't feasible with a silent shutter. So how?
I took a picture of a timer, then counted the number of frames each second. For example, at 120fps, I counted 93 frames at the 1 second mark, then 6 frames in the 2 second mark and so on. There is some margin of error but I did the tests multiple times and was consistently within a few frames.
Many thanks Rob for such an extensive comparison. It would appear that OMDS has not taken full advantage of the speed of the UHS II cards. The difference is insignificant in normal day to day shooting. Just buy the best UHS I card.
Read my comment above. Most people use the wrong cards with olympus and fuji cameras. I got 2 times better results with the transcend and kingston cards.
Thanks for the effort. It must be time consuming. My opinion about the camera and the cards - It makes no sense to make a camera that can shoot with such insane speed and limit the size of the buffer and the type of the card. It needs one of these combo slots that take CFexpress and SD cards. Especially when they aim the camera at action shooters crowd.
Very very good job. You said what settings you used exactly. It's rare. I thought about doing same test because i have two uhs2 cards, which are 2 times cheaper than sandisk and lexar, but they are much faster. They are slower or equal on the specs sheet, but Olympus cameras really like them. I got definitely better results compared to yours. Lexar cards do not work well with olympus and fuji cameras.I think this may be the cause. Prograde was a part of the lexar company. To cards, i'm talking about are kingston canvas react plus and transcend 700s. As an example i give my result for sh2 25, 50fps and sq20 with kingston and om1. Sh2 25fps- 165 images before the buffer is full, 50 fps 110-115 images before the buffer is full, clearing the buffer 8,5 and 8 seconds. 20 fps - 175 images and 8 seconds. 120fps -100 images and 7.5 seconds. 10 seconds burst 120fps-220 images and 7.5-8 seconds. Importantly, these are the results with a landscape scene. When you have less detail, like an object on a blurred background, the results are even better(about 0.5sec) Interestingly, these "professional reviewers" like Gordon or DP review tests buffers without a lens (!) with lexar or sandisk cards. They also don't say what iso they use. Raising the iso lowers results. I myself use uhs1 for video, but for photography only uhs2. Kingston sd cames with an sd reader. This reader is faster than the sandisk extreme pro reader which cost me 25 €. That's cool, but who needs five card readers. They should sold them separatly and lower the price. The transcend card is slower than the kingston, but it is even cheaper and it's faster than the old 285mb/s extreme pro sandisk, which is twice as expensive. I don't know how much these cards cost in US. Prices in Europe (no prograde cards in Europe) Sandisk 64gb v90 105 € version 2 300MB/s (slower than kingston and equal to transcend, old 280MB/s version was much slower), Lexar 2000x v90 110-120€ (the slowest of all tested uhs2 cards), Kingston canvas react plus 50-60 €, Transcend 700s v90 45€.
Thanks for sharing your results. Did you have IBIS on or off? Raw, R+J, or Jpg? As for additional settings, not sure if I mentioned I was in manual mode using fixed iso. Sometimes at 200, other times set to 1600, depending on shutter speed I needed. I took a picture of a timer and x-rite color card filling about 75% of the frame, the rest blurred out. I'll have to lookup those kingston cards you mentioned. I've heard about those being very good in Olympus cameras. For reference here in the USA, the Prograde cards cost more than the Lexar, but less than Sandisk.
@@RobTrek I turned off the ibis and set the autofocus to afs. I only tested raw and iso 200. I repeated the test with sh2 25fps with different iso values.Iso 200 165 images, cleaning buffer 7.5-8 seconds, iso 1600 140 images and 9,5 seconds,but iso 2000, 6400 and 12800 120 images and 14 seconds (big jump) I think these are good results. How about m1miii? Same story,but much worse: M1miii 18fps iso 200- 80 images, 5.5 sec, iso 1600 75images, 5 sec, but iso 2000, 6400 55 images and 14-15 sec(!) With m1miii i use electronic shutter and 10fps, sometimes 15fps. 10fps - 500 shots before the buffer is full, on higher ISOs above 1600 only 70(!), 15fps -90 and only 50 on higher ISOs. Therefore, with m1mii and miii i use max iso 1600 for shooting races. I know about kingston cards thanks to Peter Forsgard. Transcend was simply the cheapest and equaled Sandisk. Now kingston seems to be the best option. Toshiba has great cards too, but they're damn expensive. I'll do more tests with jpg, raw and dual cards later. I remember my days with the Canon R5. What's interesting is that pro reviewers say nothing about these things. For example, r5 with super expensive CF Express card on ISO 6400 has four times slower write speed compared to base ISO, with iso 1600 3200 2 times.. And r5 only has one input for cfexpress card and one for sd. Using 20fps doesn't make sense on this camera with SD cards. In addition, the quality of photos taken with the electronic shutter is much worse than compared to mechanical shutter (dynamic range, ISO and distortion) The same with sony a7. It's completely different with om1. Electronic shutter gives much better results (sharper and cleaner image). Anyway, m1miii and mii is also a little better with the electronic shutter. Overall, after playing with r5,i don't think comparing r5 and r6 to om1 makes any sense, but that's another story. Sorry for a partly offtopic post.
@@kamilrakowski23 I found the Kingston today and ordered it. Will retest and see if I get different results. Also, I've updated the numbers for dual card slots. The UHS-I takes 61 seconds to write, not 34. I do get better numbers with IBIS off. Turn IBIS on and see if your numbers are closer to mine. I may do another video on this topic to cover other scenarios and offer suggestions. As for the Canons and other brands, I'm just not familiar enough to comment. I'm am surprised the electronic shutter has lower image quality. What you said about mixing cf with sd makes sense. Thanks.
@@RobTrek I did a test with colorchecker, om1 sf2 25fps, sq 20fps, and m1miii sq18fps with three kingston cards, ibis off and on, with one card, two cards (raw + raw and raw + jpg lsf, raw + jpg lsf with noise reduction. Results and conclusions: Ibis on/off tested at 12mm, 40mm (12-40 pro) and 110mm(40-150pro) hand held. If you set the Image stabilizer to FPS priority it has no effect on the fps and buffer (for both m1miii and om1)or close to zero. I repeated the tests 4 times. Only at 110mm i got 150 pictures instead of 165 in sh2 mode, but there was no difference in fps (6 sec burst and 150fps) SQ 20fps- 8.5- 9 sec burst-170-180 pictures and 8 seconds cleaning. Another interesting thing is the difference between the cards. Sometimes I had 10 pictures more , sometimes 10 less, camera cleared the buffer in 8 seconds, sometimes in 9. I tested kingston cards in the crystal disk mark. I got 295/290, 292/290 and 292/262 MB/s. One card was slower, but still ok (specs sheet says 300/260) The other two were clearly faster than they should be. Test with 2 cards: Here are the results: two kingston cards, raw + raw. 150 images and 9.5 seconds. I have repeated this test 3 times. Each time 150 images, 9.5 sec. With iso 2000 and higher 14-15 sec. Great result im my opinion. There is one more thing I need to ask you Rob. Have you changed the settings in the green section # 3 ISO/Noise Reduction》 Low ISO Processing? You should select Drive Priority. Detail priority is by default. This setting has the greatest effect on camera speed. I think that omd should define what cards we should use and test them for us, because the results i got were clearly better than the om system gives in the specification. The second issue are reviews discouraging people from buying om1. 3-4 days ago i heard that om1 is useless for action because in sh2 25 and 50 fps mode it can only take 90 raw photos and camera clears the buffer in 55 seconds. If we use the right settings and the right cards, the results are completely different.
I had Noise Filter set to OFF and Image Stabilization Priority ON in my tests for this video, but camera forces FPS Priority in SH modes. I got my Kingston card in today! So I ran a couple of tests with IBIS OFF, Drive Priority, and Noise Filter OFF. The numbers were much closer to yours. 10 second burst SQ20 Raw 195/9 SQ20 R+J 146/16 SQ20 Jpg 204/7 SH2 Raw 195/9 SH2 R+J 145/16 SH2 Jpg 209/8 I actually got virtually identical numbers using my ProGrade cards, so that's good, but the kingstons were substantially cheaper. As you said, the Drive Priority is the difference in clearing the buffer. Total number of frames is affected by IBIS, which I mentioned in the video. I'm very surprised that Drive Priority cut all the times by more than half while capturing more images. Thanks for sharing your results. I'm going to do a followup video about this and give you credit and Peter Forsgard some credit, as Peter also mentioned it. I'm also going to recommend the Kingston Cards. I've known kingston since the computer memory days back in late 90's.
this is the exactly opposite approach to Rob's scientific measured method - when I'm out in the field and i full my uhs2 and camera starts writing into uhs1 slot, I either notice immediately the slowness in writing speed or I hit buffer quickly due to not adjusting my bursts. that is shooting in raw and comparing a 270mbs rated uhs2 to a 200mbs rated uhs1. It doesn't bother me for the most part as the idea as to use fast smaller uhs2 and offload the data to the larger slower uhs1 when there's time during the shoot. But I can imagine that comparing top of the line uhs2 to regular uhs1 may produce quite a large difference.
@@creambun1963 Unfortunately I can only buy one at a time as needed. I bought the ProGrade as it got good reviews and was kind of in the middle for price. I just picked up a Kingston brand that is cheaper and will see how it performs.
@@RobTrek Honestly, i think that for most people the cheaper 100-400 is the better option. Before spending that much money, it's a good idea to borrow both and test them. I have a 300mm f4 myself, but i use mostly two bodies at the same time. One with 40-150 pro and one with 300 f4. I was at the Danish Air Force base two weeks ago and photographed the f16 demo fighters. I wish i had 150-400 pro or even 100-400 non pro and one om1 instead of 300 f4 and 40-150, mc-14/20 and m1miii with om1. But that's just for wildlife and aviation. Motorsport, that is my true love is a different story.
@@kamilrakowski23 Yeah, the 100-400 at half the price but very close in IQ is a great option for most. I've been testing the 100-400 and 300 this past week and decided on the 300 for the f/4 aperture. A pricey decision.
I found a mistake in the spreadsheet and corrected it. Specifically, the dual card numbers were off on the UHS II + UHS I. It takes 61 seconds to clear vs 34 seconds. Sorry about that. I've updated the spreadsheet download to reflect the new numbers.
So UHS-II is about twice faster than UHS-I after all :)
Thanks!
Thank you!
Thanks! Rob, Your in depth review is stellar!
Thank you!
Holy Moly Rob! Amazing amount of work went onto this presentation, very well done. Cheers!
Thanks. It was so much work!
Hi Rob when you show the speed in which it achieves the tests covered the camera is amazing. The transfer of data from the sensor to a ready image is incredible don't you think. For me also the camera controlling focusing sensor shift to stabilise and everthing else makes the tool a work of art in itself, if I try to comprehend it.
Your effort in the work you do along with the good it provides as always is just great Rob. Thank you.
Thanks. The tech is pretty amazing when it's all put together.
You're killing it lately, Rob. Awesome info. Appreciate all of the hard work.
Thanks!
Appreciate the analysis! Interesting for most situations UHS I works just fine.
Hi Rob. I watched several of your videos yesterday, including this one. I think it was here that you mentioned that in some cases, very fast sequential shots are taken too rapidly for image stabilisation to work. If so, I just noticed in the pdf manual for my M1 ii this:
pdf page 114: cog menu: custom menu C2:
Image Stabilization
"Sets the priority function during sequential shooting.
[Fps Priority]: Shooting speed gets priority over image stabilization. The sensor will not be reset to the center during sequential shooting.
[IS Priority]: Image stabilization gets priority over shooting speed. The sensor will be reset to the center per frame of sequential shooting. The shooting speed will drop slightly."
Cheers for all your great advice - my confidence is gaining momentum!
Wow! You sure took a lot of time and effort to examine this! It is very useful for those who may already have SD cards and do not need the extra performance. Another benefit of M43! Thanks again for sharing!
Glad it was helpful!
I appreciate you also showing what settings you typically use, inside the speed comparisons and settings juxtapositions. And thanks for all the technical explanations.
Glad to help!
Great job indeed!
I am currently using an em5 miii and i shoot a lot of photos of animals in zoos. I never use very high burst modes. Because, except if your are a fan of humming birds, 10 fps are enough in main cases.
In addition, using burst mode lets you go home after three hours in a zoo with something like 1000 photos. I can t imagine if i was using 120fps.
But i must admit that in some circumstances (capture pro) it may be useful.
Anyway, i live in France (it explains my bad english) and OM-1 will be avaliable only next week, i'll buy it as soon as it will be in my favorite store.
Thanks again for your videos.
Thanks. Your english is fine!
with high burst speed you loose ability to use clusterised AF points, which important to achieve high hit AF rate with continuos AF. Be aware!
Watching at work. Well done. I appreciate your efforts.
Thanks. I try.
@@RobTrek From what I gather a 64gb card is good for around 2000 photos. Interested in your reason for 256GB cards.
This would have been just as useful as your excellent standard, had I the funds to purchase this camera.
In line with your usual practical and very high standard. Enjoyed watching, even not having the camera!
Thanks, Robert.
As I understand it, the buffer is actually emptying as fast as it can to the card and is dependent upon the speed the card can be written to, which varies from card brand/model to card. So it's not like the buffer fills up and then over flows into the memory card. It's constantly emptying to the card(s) once the shutter is fully pressed. So if a particular file format and card spec can be written to faster or as fast as the camera is recording, then conceivably, we could just shoot at that format and speed until the card's memory was full. With some cards, if recording at the extreme fps speeds beyond 20 or 30 fps, then, depending on how fast the card can be written to, the results may vary somewhat. But I found with most of our cameras, that write speed was more limited by the camera hardware than the particular card, beyond a certain threshold.
For 12MB jpegs, it's possible to have a card that can be written to at 20fps and never fill the buffer, if the card can be written to fast enough. But cards are not all equal with respect to the kind of SD card hardware and firmware a particular camera employs. I remember that the OM-D E-M1 Mark II could not write faster than 180MB/s no matter what card we used. I'm not sure what the speed for the OM-1 might be. The OM-D E-M1 Mark III was about twice as fast as the OM-D E-M1 Mark II in my tests.
Thanks for explaining part of the process that I didn't cover. There are certainly a lot of variables that affect final performance.
@@RobTrek there are a lot of variables but you did a really good job. I have done my own similar tests and your numbers look almost identical to my results. I sort of imagine that the faster the action I want to record, and the more fps I need, the shorter the duration I have to cover. The slower the action, the longer the duration but the less fps I need so it seems to work out fine and I can’t remember many times in recent years where I was hitting the buffer limits before I was done shooting.
Thank you for doing this. I have spent money on Sony Tough UHS-II cards and didn't notice any real change between them and my PNY or Angelbird UHS-I cards. But a HUGE difference is that I can shoot for longer with 256GB & 512GB UHS-I cards than I can shoot with the 64GB Sonys that I have to swap-out multiple times...(And still spend less money!) Also, is it just me, or did you actually find that the UHS-I cards offloaded FASTER than the UHS-II cards? In your timer sequences you show that the UHS-I card took 3:24.68 while the UHS-II took 4:25.19. Seems like it should be the other way-around? Faster cards SHOULD offload faster if a true UHS-II reader is used and the SSD being written to can keep up... Right? I could see if the reader only does UHS-I, and then the speed gains of UHS-II cards would be negated and even might be a little slower. Just asking, and NOT a criticism.
Ugh! I got the timings reversed. Sorry about that. I did use UHS-II reader to an NVME SSD. And yes, sometimes a UHS-II card will read slower in UHS-I card readers. My older Sandisk UHS-II card reads at about 1/2 the speed of a good UHS-I card in a UHS-1 reader. However, my newer ProGrade UHS-II cards read at the same speed as UHS-I cards in a UHS-1 reader.
Rob Trek back to doing what he does best. Teaching us how to get the best from our equipment. Superb 👌
Thanks, Rick!
Yes, awesome. Just like all the excellent videos about m10 Mark ii.
Makes the waiting time for the om-1 more and more exciting.
Thanks for this analysis! It really helps deciding whether UHSII is worth it for the average shooter.
Thanks! I've updated the spreadsheet for the dual card slot numbers. Please check that before deciding.
Wow! What a ton of work you did. I’ve been searching the internet for similar information. Thanks!
QUESTION; did you notice any difference in V30 vs. V90? I’m guessing that you didn’t based on the photo of your memory cards.
At least now you have cards for your second OM-1:)
I was going to ask the same question. I have a couple V90 and a couple v60 cards, but is it worth swiping them out of my Sony FF body? Thanks for doing all the work Rob. Very helpful
When using a single card slot, shooting Raw+Jpg, it made a big difference. From there, the differences were less noticeable, especially when using dual card slots.
I'd put the V90 in t he Sony and V60 in t he Olympus. The logic being the Sony will have bigger files to save, especially after the buffer is full.
@@RobTrek Thanks Rob
Very scientific. You covered every possibility on your spreadsheet. If I did a chart like that... it would have 2 rows and 2 columns! LOL :)
Thanks. I try.
You not only try, but always succeed. Olympus and OMS owners owe you a great debt. Have still yet to really master my Pen F, however, your excellent Pen F videos have been a great aid- manual is rather useless. Btw, whilst rarely do any post, have the Pen F set to Raw, Superfine ( thanks to your video) and jpeg. The day may come when I will get someone to do some post tweaking on an image, hence the Raw setting for each image. The downside is the time taken to read to Sd cards. I only use Scandisk. Am aware that there are now very fast ( and expensive) cards ,maybe more for the mega sized files for the new mirrorless huge files. Am willing to purchase faster reading cards, but there may be an inbuilt limit as to how quickly the camera can read to any given card, no matter how fast it's read speed .
Any tips, or links to a video would be greatly appreciated.
Thanks for testing out so many scenarios. Many reported the buffer clearing time is twice longer at ISO higher than 1600. My quick check with RAW only and single UHS-II seems to reflect that. This could explain the strange results in your spreadsheet where you had 71s for single UHS-I card but 33s for dual UHS-I at 20fps RAW+J.
My quick take is: Shooting RAW only is the only way to take advantage of UHS-II. Shooting RAW+J increases processing time so much that UHS-I is good enough.
That's interesting. All my tests were done at iso200, except one set at 1600. However, the SQ20 set was done at iso200. I just tested UHSI R+J SQ20 again and got the same result at iso200 (103 images at 71sec to clear), but did it again at iso 3200 and it took 103 images and 83 seconds to clear. So it seems iso processing is putting a drag on the write times. Thanks. Also, you've found a mistake I made in the spreadsheet and I've updated it. It takes 61 seconds to clear when using dual card slots with UHS-I in one or both slots.
Hi Rob, very elaborate and certainly useful. I don't think that shooting for 10sec is very realistic, at least for adventure, sports or wildlife, which I shoot regularly. Horse in races, run avg. 50Km/h (~30mph), so in 1 sec horses have moved ~14m (46ft) and a burst of 1 to 2 sec, possibly panning, is actually long. Using ProCap with E-M1 III and now OM-1, I find that 20-50 frames are more than adequate to capture the moment. Need less frames if using super telephoto like 300/4 Pro to keep subject in the frame. Need more frames using wider view to capture butterfly taking off with 40-150/2.8 Pro. Capturing more frames is probably adding frames of little, if any, use. Therefore, IMHO the buffers of E-M1 III, E-M1X and OM-1 are in practice more than adequate. With this thinking, UHS-II cards with the highest write about 250-280Mbps are very useful to empty the buffer quickly, These cameras actually allow to capture a new sequence before the buffer is empty, so speed is even more useful.
Thanks for sharing your experience. I'm finding that the buffer is sufficient in real world usage as well.
Thank you for this research. I thought it would make a bigger difference. This is helpfull, saving monney!
Glad it was helpful!
How did you measure the buffer speed. The buffer is an internal camera memory and the read/write speed should be always the same. When the buffer is full then latest the writing on SD card starts. And writing speed of SD cards differs a lot by specification, though also influenced by camera interface.
The buffer seems to capture 93fps. I took a picture of a timer, then counted the number of frames each second. For example, at 120fps, I counted 93 frames at the 1 second mark, then 6 frames in the 2 second mark and so on. There is some margin of error but I did the tests multiple times and was consistently within a few frames. Near the end of the the video, I showed the cards I used and a standard crystal diskmark score.
@@RobTrek I find it varies a little bit depending on the card since the buffer is emptying continuously once the shutter is fully pressed so if a particular card can be written to faster than another card, the results will reflect that.
Thank you ! Very useful information. I have two cards: uhs I and uhs II, If I only shoot raw and use one card at each time, so my best choice is to save to the uhs II card first, am I right? Thank you very much!
Yes!
Fast cards are worth the money if you are a burst shooter. The upload speed is just a bonus. Thanks for the data validation. I didn't need to be sold, but am certainly vindicated in springing for the extra cost of fast cards.
Definitely! I'm going to do a followup video on this. I want to explore how different settings affect performance. It's pretty substantial.
Nice details thanks, unfortunately you did not state the speed of the cards ush1 and 2 does not indicate the speed of the cards ie v60 v90
Sorry, I thought I did. UHS-II V90 and UHS-I V30.
great stuff thanks for all the info
Thanks!
Thank you for bringing us this information. What struck me is that UHS-I and UHS-II are almost identical. I tried to mimic your test wth my SD cards and I get about the same performance. What I didn't figure out is how you find the FPS value with buffer full. I used to record the shutter sound but that isn't feasible with a silent shutter. So how?
I took a picture of a timer, then counted the number of frames each second. For example, at 120fps, I counted 93 frames at the 1 second mark, then 6 frames in the 2 second mark and so on. There is some margin of error but I did the tests multiple times and was consistently within a few frames.
Many thanks Rob for such an extensive comparison. It would appear that OMDS has not taken full advantage of the speed of the UHS II cards. The difference is insignificant in normal day to day shooting. Just buy the best UHS I card.
Read my comment above. Most people use the wrong cards with olympus and fuji cameras. I got 2 times better results with the transcend and kingston cards.
Thanks! Yes, I'm returning one of my UHS II cards. I should return both but sometimes the UHS II is better.
Thanks for the effort. It must be time consuming. My opinion about the camera and the cards - It makes no sense to make a camera that can shoot with such insane speed and limit the size of the buffer and the type of the card. It needs one of these combo slots that take CFexpress and SD cards. Especially when they aim the camera at action shooters crowd.
Thanks. Agreed. A CFexpress would definitely take the camera up a notch.
Very very good job. You said what settings you used exactly. It's rare. I thought about doing same test because i have two uhs2 cards, which are 2 times cheaper than sandisk and lexar, but they are much faster. They are slower or equal on the specs sheet, but Olympus cameras really like them.
I got definitely better results compared to yours. Lexar cards do not work well with olympus and fuji cameras.I think this may be the cause. Prograde was a part of the lexar company.
To cards, i'm talking about are kingston canvas react plus and transcend 700s. As an example i give my result for sh2 25, 50fps and sq20 with kingston and om1. Sh2 25fps- 165 images before the buffer is full, 50 fps 110-115 images before the buffer is full, clearing the buffer 8,5 and 8 seconds. 20 fps - 175 images and 8 seconds. 120fps -100 images and 7.5 seconds. 10 seconds burst 120fps-220 images and 7.5-8 seconds. Importantly, these are the results with a landscape scene. When you have less detail, like an object on a blurred background, the results are even better(about 0.5sec) Interestingly, these "professional reviewers" like Gordon or DP review tests buffers without a lens (!) with lexar or sandisk cards. They also don't say what iso they use. Raising the iso lowers results. I myself use uhs1 for video, but for photography only uhs2. Kingston sd cames with an sd reader. This reader is faster than the sandisk extreme pro reader which cost me 25 €. That's cool, but who needs five card readers. They should sold them separatly and lower the price. The transcend card is slower than the kingston, but it is even cheaper and it's faster than the old 285mb/s extreme pro sandisk, which is twice as expensive.
I don't know how much these cards cost in US. Prices in Europe (no prograde cards in Europe) Sandisk 64gb v90 105 € version 2 300MB/s (slower than kingston and equal to transcend, old 280MB/s version was much slower), Lexar 2000x v90 110-120€ (the slowest of all tested uhs2 cards), Kingston canvas react plus 50-60 €, Transcend 700s v90 45€.
Thanks for sharing your results. Did you have IBIS on or off? Raw, R+J, or Jpg? As for additional settings, not sure if I mentioned I was in manual mode using fixed iso. Sometimes at 200, other times set to 1600, depending on shutter speed I needed. I took a picture of a timer and x-rite color card filling about 75% of the frame, the rest blurred out. I'll have to lookup those kingston cards you mentioned. I've heard about those being very good in Olympus cameras. For reference here in the USA, the Prograde cards cost more than the Lexar, but less than Sandisk.
@@RobTrek I turned off the ibis and set the autofocus to afs. I only tested raw and iso 200. I repeated the test with sh2 25fps with different iso values.Iso 200 165 images, cleaning buffer 7.5-8 seconds, iso 1600 140 images and 9,5 seconds,but iso 2000, 6400 and 12800 120 images and 14 seconds (big jump) I think these are good results. How about m1miii? Same story,but much worse:
M1miii 18fps iso 200- 80 images, 5.5 sec, iso 1600 75images, 5 sec, but iso 2000, 6400 55 images and 14-15 sec(!)
With m1miii i use electronic shutter and 10fps, sometimes 15fps. 10fps - 500 shots before the buffer is full, on higher ISOs above 1600 only 70(!), 15fps -90 and only 50 on higher ISOs. Therefore, with m1mii and miii i use max iso 1600 for shooting races.
I know about kingston cards thanks to Peter Forsgard. Transcend was simply the cheapest and equaled Sandisk. Now kingston seems to be the best option. Toshiba has great cards too, but they're damn expensive. I'll do more tests with jpg, raw and dual cards later.
I remember my days with the Canon R5. What's interesting is that pro reviewers say nothing about these things. For example, r5 with super expensive CF Express card on ISO 6400 has four times slower write speed compared to base ISO, with iso 1600 3200 2 times.. And r5 only has one input for cfexpress card and one for sd. Using 20fps doesn't make sense on this camera with SD cards. In addition, the quality of photos taken with the electronic shutter is much worse than compared to mechanical shutter (dynamic range, ISO and distortion) The same with sony a7. It's completely different with om1. Electronic shutter gives much better results (sharper and cleaner image). Anyway, m1miii and mii is also a little better with the electronic shutter. Overall, after playing with r5,i don't think comparing r5 and r6 to om1 makes any sense, but that's another story. Sorry for a partly offtopic post.
@@kamilrakowski23 I found the Kingston today and ordered it. Will retest and see if I get different results. Also, I've updated the numbers for dual card slots. The UHS-I takes 61 seconds to write, not 34. I do get better numbers with IBIS off. Turn IBIS on and see if your numbers are closer to mine. I may do another video on this topic to cover other scenarios and offer suggestions. As for the Canons and other brands, I'm just not familiar enough to comment. I'm am surprised the electronic shutter has lower image quality. What you said about mixing cf with sd makes sense. Thanks.
@@RobTrek I did a test with colorchecker, om1 sf2 25fps, sq 20fps, and m1miii sq18fps with three kingston cards, ibis off and on, with one card, two cards (raw + raw and raw + jpg lsf, raw + jpg lsf with noise reduction. Results and conclusions: Ibis on/off tested at 12mm, 40mm (12-40 pro) and 110mm(40-150pro) hand held. If you set the Image stabilizer to FPS priority it has no effect on the fps and buffer (for both m1miii and om1)or close to zero. I repeated the tests 4 times. Only at 110mm i got 150 pictures instead of 165 in sh2 mode, but there was no difference in fps (6 sec burst and 150fps) SQ 20fps- 8.5- 9 sec burst-170-180 pictures and 8 seconds cleaning.
Another interesting thing is the difference between the cards. Sometimes I had 10 pictures more , sometimes 10 less, camera cleared the buffer in 8 seconds, sometimes in 9. I tested kingston cards in the crystal disk mark. I got 295/290, 292/290 and 292/262 MB/s. One card was slower, but still ok (specs sheet says 300/260) The other two were clearly faster than they should be.
Test with 2 cards: Here are the results: two kingston cards, raw + raw. 150 images and 9.5 seconds. I have repeated this test 3 times. Each time 150 images, 9.5 sec. With iso 2000 and higher 14-15 sec. Great result im my opinion.
There is one more thing I need to ask you Rob. Have you changed the settings in the green section # 3 ISO/Noise Reduction》 Low ISO Processing? You should select Drive Priority. Detail priority is by default. This setting has the greatest effect on camera speed.
I think that omd should define what cards we should use and test them for us, because the results i got were clearly better than the om system gives in the specification. The second issue are reviews discouraging people from buying om1. 3-4 days ago i heard that om1 is useless for action because in sh2 25 and 50 fps mode it can only take 90 raw photos and camera clears the buffer in 55 seconds. If we use the right settings and the right cards, the results are completely different.
I had Noise Filter set to OFF and Image Stabilization Priority ON in my tests for this video, but camera forces FPS Priority in SH modes. I got my Kingston card in today! So I ran a couple of tests with IBIS OFF, Drive Priority, and Noise Filter OFF. The numbers were much closer to yours. 10 second burst
SQ20 Raw 195/9
SQ20 R+J 146/16
SQ20 Jpg 204/7
SH2 Raw 195/9
SH2 R+J 145/16
SH2 Jpg 209/8
I actually got virtually identical numbers using my ProGrade cards, so that's good, but the kingstons were substantially cheaper.
As you said, the Drive Priority is the difference in clearing the buffer. Total number of frames is affected by IBIS, which I mentioned in the video. I'm very surprised that Drive Priority cut all the times by more than half while capturing more images.
Thanks for sharing your results. I'm going to do a followup video about this and give you credit and Peter Forsgard some credit, as Peter also mentioned it. I'm also going to recommend the Kingston Cards. I've known kingston since the computer memory days back in late 90's.
this is the exactly opposite approach to Rob's scientific measured method - when I'm out in the field and i full my uhs2 and camera starts writing into uhs1 slot, I either notice immediately the slowness in writing speed or I hit buffer quickly due to not adjusting my bursts. that is shooting in raw and comparing a 270mbs rated uhs2 to a 200mbs rated uhs1. It doesn't bother me for the most part as the idea as to use fast smaller uhs2 and offload the data to the larger slower uhs1 when there's time during the shoot. But I can imagine that comparing top of the line uhs2 to regular uhs1 may produce quite a large difference.
Thanks for sharing your experience.
That 99 photo buffer sure makes a difference.
It really helped capturing more shots over the EM1.3.
can you test for the fastest sd cards for the om1 please
I've ordered a new card that should be a little faster. I doubt it will make much difference but i'll update everyone after I test. Thanks.
@@RobTrek just wondered the best budget uhs2 card vs the expensive ones would be good feature
@@creambun1963 Unfortunately I can only buy one at a time as needed. I bought the ProGrade as it got good reviews and was kind of in the middle for price. I just picked up a Kingston brand that is cheaper and will see how it performs.
@@RobTrek get um to send some free ones out for reviews and testing
@@creambun1963 I've never been offered free cards. Maybe I need to do a video reviewing cards to get other manufacturers attention.
All this OM1 buffer talk and i am still trying to justify buying the OM 20mm. 🐢
I'm trying to justify the 300mm f/4 myself. Expensive hobby!
@@RobTrek I have the 300 my advice get it
@@RobTrek Honestly, i think that for most people the cheaper 100-400 is the better option. Before spending that much money, it's a good idea to borrow both and test them. I have a 300mm f4 myself, but i use mostly two bodies at the same time. One with 40-150 pro and one with 300 f4. I was at the Danish Air Force base two weeks ago and photographed the f16 demo fighters. I wish i had 150-400 pro or even 100-400 non pro and one om1 instead of 300 f4 and 40-150, mc-14/20 and m1miii with om1. But that's just for wildlife and aviation. Motorsport, that is my true love is a different story.
@@kamilrakowski23 Yeah, the 100-400 at half the price but very close in IQ is a great option for most. I've been testing the 100-400 and 300 this past week and decided on the 300 for the f/4 aperture. A pricey decision.
I have both, the keeper rate with the 300 is so much higher it's ridiculous. Get the 300.