macgregor
Veteran
Bruce Allen and myself spent some time playing with the Samsung.
Here you can find an in depth review: http://www.migueldeolaso.com/press/
Here you can find an in depth review: http://www.migueldeolaso.com/press/
Samsung NX1 - all sorts of tests imaginable
- Thread starter macgregor
- Start date
Bassman2003
Veteran
Thanks for posting. Pretty impressive. Amazing how close the NX1 and the Sony are with color and overall look.
diegocervo
Well-known member
hey Mac,
thanks for the time you took for this tests.
How did you convert h265 footage? Usually I don't get those compression artifacts when using Rocky Mountain, converting files to H264 first.
P.s. Are you in Panama City? We can meet for a beer
thanks for the time you took for this tests.
How did you convert h265 footage? Usually I don't get those compression artifacts when using Rocky Mountain, converting files to H264 first.
P.s. Are you in Panama City? We can meet for a beer
macgregor
Veteran
Well... this is what I posted in my site:
The NX1 seems to produce slightly sharper images at 4K. But it introduces very visible sharpening artifacts. The A7s doesn’t seem to resolve the same amount of detail. Of course this difference is totally invisible when watching the footage at 1080p and even in 4K the differences are subtle.
I don't care that much about resolution, colorimetry and DR are more interesting to me. And price. The A7s in 4K costs 3 to 4 times what a NX1 costs. How can we compare these two? That means I can buy two Samsung bodies, and one 16-50 and one 50-150mm lenses, and have a super compact guerrilla style kit. With the same money all I get is one A7s with an odyssey but no lenses, so at the end of the day the cost is completely different.
I like DSLR that record internally. If you need to carry a monitor/recorder, that requires extra batteries and stupid tiny hdmi cables... that is no longer guerrilla. And for the cost you might as well just jump over an FS7 which will be infinitely better. I guess this is personal, but to me the only advantage to DSLRs this days are size and being consider a "tourist". No extra cables, no complicated rigs. SImplest camera set up ever.
Yes I would take an A7s+Odyssey for a serious narrative project over the NX1 at gunpoint. I actually own 2 A7s so far. But the A7s needs batteries every 30 minutes, external recorder balblah... not worth it.
The Samsung has a problem with the sharpening though. They better fix that or the footage is kind of useless in 4K mode.
One thing I forgot to mention is that while the NX1 seems to be only capable of 10 stops of DR in video mode, the rollof in the highlights is extremely soft (as opposed to cameras like Nikon or the A7s in cine4 gamma mode that go from detail to blown out highlights abruptly). This is actually pretty good.
The NX1 seems to produce slightly sharper images at 4K. But it introduces very visible sharpening artifacts. The A7s doesn’t seem to resolve the same amount of detail. Of course this difference is totally invisible when watching the footage at 1080p and even in 4K the differences are subtle.
I don't care that much about resolution, colorimetry and DR are more interesting to me. And price. The A7s in 4K costs 3 to 4 times what a NX1 costs. How can we compare these two? That means I can buy two Samsung bodies, and one 16-50 and one 50-150mm lenses, and have a super compact guerrilla style kit. With the same money all I get is one A7s with an odyssey but no lenses, so at the end of the day the cost is completely different.
I like DSLR that record internally. If you need to carry a monitor/recorder, that requires extra batteries and stupid tiny hdmi cables... that is no longer guerrilla. And for the cost you might as well just jump over an FS7 which will be infinitely better. I guess this is personal, but to me the only advantage to DSLRs this days are size and being consider a "tourist". No extra cables, no complicated rigs. SImplest camera set up ever.
Yes I would take an A7s+Odyssey for a serious narrative project over the NX1 at gunpoint. I actually own 2 A7s so far. But the A7s needs batteries every 30 minutes, external recorder balblah... not worth it.
The Samsung has a problem with the sharpening though. They better fix that or the footage is kind of useless in 4K mode.
One thing I forgot to mention is that while the NX1 seems to be only capable of 10 stops of DR in video mode, the rollof in the highlights is extremely soft (as opposed to cameras like Nikon or the A7s in cine4 gamma mode that go from detail to blown out highlights abruptly). This is actually pretty good.
Last edited:
macgregor
Veteran
hey Mac,
thanks for the time you took for this tests.
How did you convert h265 footage? Usually I don't get those compression artifacts when using Rocky Mountain, converting files to H264 first.
P.s. Are you in Panama City? We can meet for a beer
I did use rocky mountain but trasncoded everything to Prores HQ.
How long are you in Panama for?
I go there very often. Send me a PM!
Samuel H
Major Contributor
Thanks a lot macgregor, I really enjoyed reading all your NX1 thought on your website.
The nice highlight rolloff (on the a7S the sky whold be showing lots of funky blue color bands in the transition to the clipping point) and the nasty sharpening hadn't been highlighted enough, I think.
The DR thing is what bugs me most. I find myself struggling with the a7S+slog2 on some occasions already, I wouldn't want to go "two stops less", even with a nicer highlight rolloff. Let's see if Samsung can listen and implement a wider gamma curve. According to dxomark and my a7S stills-vs-video tests here, the NX1 sensor has the capability to provide at least 1.5 stops more DR than what we are getting with the a7S+slog2 in video mode.
And I completely share your thought that external 4K on the a7S is cumbersome and expensive. I decided to not buy an external recorder, and wait for new and better cameras instead. This is very personal, but a bigger camera package doesn't match *my* shooting style.
The nice highlight rolloff (on the a7S the sky whold be showing lots of funky blue color bands in the transition to the clipping point) and the nasty sharpening hadn't been highlighted enough, I think.
The DR thing is what bugs me most. I find myself struggling with the a7S+slog2 on some occasions already, I wouldn't want to go "two stops less", even with a nicer highlight rolloff. Let's see if Samsung can listen and implement a wider gamma curve. According to dxomark and my a7S stills-vs-video tests here, the NX1 sensor has the capability to provide at least 1.5 stops more DR than what we are getting with the a7S+slog2 in video mode.
And I completely share your thought that external 4K on the a7S is cumbersome and expensive. I decided to not buy an external recorder, and wait for new and better cameras instead. This is very personal, but a bigger camera package doesn't match *my* shooting style.
diegocervo
Well-known member
Apparently downgrading is possible if you still have firmware 1.2 stored somewhere.
http://www.dpreview.com/forums/thread/3847445
Next time, please try H265 to H264 in RM, then H264 to Prores HQ in FCPX.
Time consuming, but worth the extra coffee break eheheh
Cheers,
Diego
http://www.dpreview.com/forums/thread/3847445
Next time, please try H265 to H264 in RM, then H264 to Prores HQ in FCPX.
Time consuming, but worth the extra coffee break eheheh
Cheers,
Diego
Samuel H
Major Contributor
Why do you need to go from H264 to ProRes? Can't you just recode from H265 to H264 and then edit that? With modern editing software you're not going to have any improvements in terms of color depth and such, and I thought a half-decent current workstation would be enough to work with 4K H464.
diegocervo
Well-known member
Well, my macbook pro retina really struggles in DaVinci with 4k h264, even when working on an HD timeline or with proxy enabled.
As I said, converting to prores may add another step to this workflow, but FCPX does it very quickly and my machine works faster when editing and grading. Overall it works for me.
I'm sure that a beefier computer can handle 4k H264 much better at all stages, but I'm waiting to upgrade my system.
Now I'm just investing in lenses, gymbal and lights lights lights
As I said, converting to prores may add another step to this workflow, but FCPX does it very quickly and my machine works faster when editing and grading. Overall it works for me.
I'm sure that a beefier computer can handle 4k H264 much better at all stages, but I'm waiting to upgrade my system.
Now I'm just investing in lenses, gymbal and lights lights lights
reem12
Veteran
I completely agree with diegocervo. The Rm transcode to h264, then to prores has rendered my footage very clean and artifact free. Actually cleaner than my old a7s footage. Going directly to prores has given me artifacts almost every time. to sure why, but I avoid that at all cost.
macgregor
Veteran
Does anyone have it?
improser
Active member
Very thanks for the review.
I have some questions if possible:
1. Overall in internal recording, at 1.20 firm, the C4K were softer than UHD mode. In the still from DPReview of video on charts this is very visible, and commented in other places. This change in 1.30?
2. What framerates support in C4K and UHD over HDMI output. I read somewere that only works 30fps but not the others.
3. Sometimes you test the sharpening at 0 and others at -10, but aparently even at -10 it shows artifacts, but its not clear to me.
4. Can you post somewhere the original files from camera and the Odyssey to inspect?
Im very grateful and confident of your review, I espect this type on info to finally buy one.
As a side note, I created a community open document to summarize the bugs and other stuff that dont work where people suggest and complete for a request: https://drive.google.com/open?id=1pwCklH7KMPvGlOWiy9RjGKUKaQnD7gJ-hiW7xotVJ6o&authuser=0
I have some questions if possible:
1. Overall in internal recording, at 1.20 firm, the C4K were softer than UHD mode. In the still from DPReview of video on charts this is very visible, and commented in other places. This change in 1.30?
2. What framerates support in C4K and UHD over HDMI output. I read somewere that only works 30fps but not the others.
3. Sometimes you test the sharpening at 0 and others at -10, but aparently even at -10 it shows artifacts, but its not clear to me.
4. Can you post somewhere the original files from camera and the Odyssey to inspect?
Im very grateful and confident of your review, I espect this type on info to finally buy one.
As a side note, I created a community open document to summarize the bugs and other stuff that dont work where people suggest and complete for a request: https://drive.google.com/open?id=1pwCklH7KMPvGlOWiy9RjGKUKaQnD7gJ-hiW7xotVJ6o&authuser=0
macgregor
Veteran
1. Overall in internal recording, at 1.20 firm, the C4K were softer than UHD mode. In the still from DPReview of video on charts this is very visible, and commented in other places. This change in 1.30? On 1.30 the detail resolved is pretty much the same. I did look carefully about this and they seem to deliver the same image. The 4K might be an upresized UHD though, as I see some artifacts that could be product of a quick upscale.
2. What framerates support in C4K and UHD over HDMI output. I read somewere that only works 30fps but not the others. In NTSC I only got 30p (fm 1.20), in PAL you get 25fps. I haven't tested fm1.3
3. Sometimes you test the sharpening at 0 and others at -10, but aparently even at -10 it shows artifacts, but its not clear to me. Correct. I left sharpening all the way down as I did more and more testing. It's the biggest downside of this camera. They should switch it off.
4. Can you post somewhere the original files from camera and the Odyssey to inspect? I'm afraid they're too big, sorry.
2. What framerates support in C4K and UHD over HDMI output. I read somewere that only works 30fps but not the others. In NTSC I only got 30p (fm 1.20), in PAL you get 25fps. I haven't tested fm1.3
3. Sometimes you test the sharpening at 0 and others at -10, but aparently even at -10 it shows artifacts, but its not clear to me. Correct. I left sharpening all the way down as I did more and more testing. It's the biggest downside of this camera. They should switch it off.
4. Can you post somewhere the original files from camera and the Odyssey to inspect? I'm afraid they're too big, sorry.
macgregor
Veteran
here we go!
I was able to downgrade to v1.20 and do a test comparison vs v1.30. Aaaaand there is a clear difference! (v1.20 on top, v1.30 bottom). This was quite a heavy contrast scene.
Shot with the same nikon 14-24 Nikon lens. Same exact settings: UHD 24p, internal recording, DR gamma, sharpen -10, black levels +4. Daylight WB. ISO 200.
The newer v1.30 seems to apply a different kind of sharpening (not sure if more or less, just different, check the STOP sign). BUT most important, the gamma curve has changed (less contrast)!!!
I think the new upgrade might have at least one more stop in the shadows.
I was able to downgrade to v1.20 and do a test comparison vs v1.30. Aaaaand there is a clear difference! (v1.20 on top, v1.30 bottom). This was quite a heavy contrast scene.
Shot with the same nikon 14-24 Nikon lens. Same exact settings: UHD 24p, internal recording, DR gamma, sharpen -10, black levels +4. Daylight WB. ISO 200.
The newer v1.30 seems to apply a different kind of sharpening (not sure if more or less, just different, check the STOP sign). BUT most important, the gamma curve has changed (less contrast)!!!
I think the new upgrade might have at least one more stop in the shadows.
Last edited:
I've used Photon (http://brightland.com/w/photon-manual/) for converting the footage using ProRes Mode 2, love the results. What I think is being understated in this update is the great improvement to 1080p, incredible quality now from it. Still hoping for H265 support in the next premiere update, really want to edit the native files.
Last edited:
improser
Active member
@macgregor What you thinks about the NX1 vs. GH4. I sold my GH4 for a time, but when I think to repurchase (having a SpeedBooster waiting for it) the NX1 appears so Im hold the purchase. The GH4 is very refined and knowed by me, but apparently the NX1 has better tech in all ways, but the ****ing current implementation of encoder or the bitrate asigned that manifiest in that blocking artifacts in every pseudo plain surface or low light zone.
P.D.: sorry for my english.
P.D.: sorry for my english.