PDA

View Full Version : High ISO and Panasonic S1r vs S1



ron787
8th January 2020, 10:24
I am a Panasonic S1r owner/user and am aware of the stated superiority of the S1 vs S1r when it comes to low light/high ISO situations. And there are several Internet posts that attempt to demonstrate the difference via their low quality photos (low quality due to Internet requirements), but I cannot find any information regarding how many stops of actual improvement the S1 offers when compared to the S1r?

I have not been disappointed with the S1r's low light capability, but if the S1 offers a quantifiably, significant low light advantage I might consider adding it to my stable of camera bodies.

Thank you in advance,
Ron

iiiNelson
8th January 2020, 19:02
I am a Panasonic S1r owner/user and am aware of the stated superiority of the S1 vs S1r when it comes to low light/high ISO situations. And there are several Internet posts that attempt to demonstrate the difference via their low quality photos (low quality due to Internet requirements), but I cannot find any information regarding how many stops of actual improvement the S1 offers when compared to the S1r?

I have not been disappointed with the S1r's low light capability, but if the S1 offers a quantifiably, significant low light advantage I might consider adding it to my stable of camera bodies.

Thank you in advance,
Ron

By most accounts it gives a 1-2 stop advantage (in favor of the S1) depending on the scene/subject.

ron787
9th January 2020, 06:54
By most accounts it gives a 1-2 stop advantage (in favor of the S1) depending on the scene/subject.

Thank you, your information is helpful.
One to two stops under accommodating conditions seems hardly enough to justify the addition of the S1 to my S1r ownership. Since I do not typically shoot in caves or dimly lit nightclubs, should the necessity arise I will resort to a fast lens, such as the Sigma 35mm L F/1.2, and deal with whatever noise in post.

That said, I had pushed the S1r to ISO 6400 on one occasion in a room lit by a single LED bulb and had not found the minimal noise to be objectionable.

iiiNelson
9th January 2020, 10:03
Thank you, your information is helpful.
One to two stops under accommodating conditions seems hardly enough to justify the addition of the S1 to my S1r ownership. Since I do not typically shoot in caves or dimly lit nightclubs, should the necessity arise I will resort to a fast lens, such as the Sigma 35mm L F/1.2, and deal with whatever noise in post.

That said, I had pushed the S1r to ISO 6400 on one occasion in a room lit by a single LED bulb and had not found the minimal noise to be objectionable.

From most samples Iíve seen the biggest advantages are when shooting ISO 6400-51200. IMO the ISO 6400 on the S1R looks like ISO 25600 on the S1 and 12800 on S1R looks slightly worse than 51200 on S1. I think 12800 is very useable on the S1R for B&W shots and SOME color shots depending on delivery platform. Iím looking into adding a S1H for video (if I decide to add another body) but itís largely the same as the S1 with added video specs and an AA filter.

SrMphoto
9th January 2020, 10:30
From most samples I’ve seen the biggest advantages are when shooting ISO 6400-51200. IMO the ISO 6400 on the S1R looks like ISO 25600 on the S1 and 12800 on S1R looks slightly worse than 51200 on S1. I think 12800 is very useable on the S1R for B&W shots and SOME color shots depending on delivery platform. I’m looking into adding a S1H for video (if I decide to add another body) but it’s largely the same as the S1 with added video specs and an AA filter.

Is your noise observation at pixel level or when scaling down S1R image to the same size as S1 image?

ron787
9th January 2020, 11:50
From most samples Iíve seen the biggest advantages are when shooting ISO 6400-51200. IMO the ISO 6400 on the S1R looks like ISO 25600 on the S1 and 12800 on S1R looks slightly worse than 51200 on S1. I think 12800 is very useable on the S1R for B&W shots and SOME color shots depending on delivery platform. Iím looking into adding a S1H for video (if I decide to add another body) but itís largely the same as the S1 with added video specs and an AA filter.

Thank you.

iiiNelson
9th January 2020, 12:52
Thank you.

No scaling involved for that difference. I imagine that scaling might reduce the difference quite a bit... maybe. I don’t have the S1 to try myself and it seems most online review sources (that aren’t DPReview) are of one and not the other. Unfortunately most S1 reviews lean more heavily into the video side of things because so many wanted it to be the FF GH camera that the S1H is despite Panasonic repeatedly saying otherwise.

SrMphoto
9th January 2020, 13:11
No scaling involved for that difference. I imagine that scaling might reduce the difference quite a bit... maybe. I donít have the S1 to try myself and it seems most online review sources (that arenít DPReview) are of one and not the other. Unfortunately most S1 reviews lean more heavily into the video side of things because so many wanted it to be the FF GH camera that the S1H is despite Panasonic repeatedly saying otherwise.

It does not make sense to compare noise at different image dimensions.

The question was should the OP get an S1 or stay with S1R. If an S1R image that is scaled down to the dimensions of an S1 image has approximately the same noise (it does not, but is not that far away), then it would not make much sense to get S1 because of noise characteristics.

iiiNelson
9th January 2020, 16:16
It does not make sense to compare noise at different image dimensions.

The question was should the OP get an S1 or stay with S1R. If an S1R image that is scaled down to the dimensions of an S1 image has approximately the same noise (it does not, but is not that far away), then it would not make much sense to get S1 because of noise characteristics.

Fair enough. Iím comparing results from Panasonic press events because thatís the only place I can consistently find reviews other than DPReview that test the noise between the two cameras. In every case the S1 has a significant advantage of a stop or two depending on the scene.

SrMphoto
9th January 2020, 16:38
Fair enough. Iím comparing results from Panasonic press events because thatís the only place I can consistently find reviews other than DPReview that test the noise between the two cameras. In every case the S1 has a significant advantage of a stop or two depending on the scene.

Maybe DPR's studio scene (https://www.dpreview.com/reviews/image-comparison/fullscreen?attr18=daylight&attr13_0=panasonic_dcs1&attr13_1=panasonic_dcs1r&attr13_2=leica_sl_typ601&attr13_3=leica_q2&attr15_0=jpeg&attr15_1=jpeg&attr15_2=jpeg&attr15_3=jpeg&attr16_0=6400&attr16_1=6400&attr16_2=6400&attr16_3=6400&attr126_0=1&attr126_1=1&attr126_2=1&attr126_3=1&normalization=compare&widget=725&x=0.5571599794532274&y=-0.9755785159251984) is the best available way to compare noise.

ron787
10th January 2020, 06:58
Maybe DPR's studio scene (https://www.dpreview.com/reviews/image-comparison/fullscreen?attr18=daylight&attr13_0=panasonic_dcs1&attr13_1=panasonic_dcs1r&attr13_2=leica_sl_typ601&attr13_3=leica_q2&attr15_0=jpeg&attr15_1=jpeg&attr15_2=jpeg&attr15_3=jpeg&attr16_0=6400&attr16_1=6400&attr16_2=6400&attr16_3=6400&attr126_0=1&attr126_1=1&attr126_2=1&attr126_3=1&normalization=compare&widget=725&x=0.5571599794532274&y=-0.9755785159251984) is the best available way to compare noise.

I've reviewed DPR's S1/r comparison, and to my eyes, up to ISO 25600, the S1r images appears smoother and less noisy. Of course, these are jpg images reduced for the Net. A better comparison might be accomplished with raw images, where in-camera noise reduction has presumably not taken place.
So, are my eyes fooling me, or is the DPR comparison flawed?

SrMphoto
10th January 2020, 18:07
I've reviewed DPR's S1/r comparison, and to my eyes, up to ISO 25600, the S1r images appears smoother and less noisy. Of course, these are jpg images reduced for the Net. A better comparison might be accomplished with raw images, where in-camera noise reduction has presumably not taken place.
So, are my eyes fooling me, or is the DPR comparison flawed?


My mistake, I gave you the link to JPEG comparison. Here is the link for RAW file comparisons (https://www.dpreview.com/reviews/image-comparison/fullscreen?attr18=daylight&attr13_0=panasonic_dcs1&attr13_1=panasonic_dcs1r&attr13_2=leica_sl_typ601&attr13_3=leica_q2&attr15_0=raw&attr15_1=raw&attr15_2=raw&attr15_3=raw&attr16_0=6400&attr16_1=6400&attr16_2=6400&attr16_3=6400&attr126_0=1&attr126_1=1&attr126_2=1&attr126_3=1&normalization=compare&widget=725&x=0.42874265167513254&y=0.8816497310221161) (you can easily switch between JPEG and RAW comparison on the studio scene yourself). One can also download the RAW files and test them in the preferred software.

Indeed, the JPEG files look better with S1R. Maybe that means that with S1R files we can do better noise reduction than with S1 files?

ron787
11th January 2020, 08:10
My mistake, I gave you the link to JPEG comparison. Here is the link for RAW file comparisons (https://www.dpreview.com/reviews/image-comparison/fullscreen?attr18=daylight&attr13_0=panasonic_dcs1&attr13_1=panasonic_dcs1r&attr13_2=leica_sl_typ601&attr13_3=leica_q2&attr15_0=raw&attr15_1=raw&attr15_2=raw&attr15_3=raw&attr16_0=6400&attr16_1=6400&attr16_2=6400&attr16_3=6400&attr126_0=1&attr126_1=1&attr126_2=1&attr126_3=1&normalization=compare&widget=725&x=0.42874265167513254&y=0.8816497310221161) (you can easily switch between JPEG and RAW comparison on the studio scene yourself). One can also download the RAW files and test them in the preferred software.

Indeed, the JPEG files look better with S1R. Maybe that means that with S1R files we can do better noise reduction than with S1 files?

Thank you for the RAW file link.
I've downloaded the comparative raw files from between ISO 6400 to 25,000 and opened them in Capture One Pro 20. The most noticeable difference—to my eyes and with my calibrated monitor—appears between 12,000 (where 12,000 might be usable from both cameras. s1 more so than S1r) and 25,000, with the least noise apparent in the S1 samples. However, while there is a difference, it is not as significant as I'd anticipated, and I wonder—for myself—how often I'd be required to use ISO 12,000 or for that matter, 25,000.
When faced with a photographic opportunity that knowingly requires added light, I would bring along a flash and a fast lens. But, admittedly, there are venues where such artificial light sources might be distracting and or disallowed, and in that case C!'s noise reduction feature might come in handy for the S1r file.
I'm still on the fence in re the addition of an S1 to my S1r, I simply have to determine whether the very high ISO differential is worth the loss of detail afforded by the large sensor.
Thanks again.
Ron

P.S. I've just completed a comparison of the S1raw ISO 25000 file viewed at 100% in C1against the S1r raw file at 100% with C1's noise reduction set to #30 and aside from the fact that the S1r's image is larger, they reveal the same degree of limited noise as well as the same degree of sharpness (noise reduction tends to reduce sharpness to some degree depending upon the intensity employed). There is no visible color alteration. If this approach were to hold true for images acquired under other circumstances the S1 vs S1r High ISO differential up to ISO 25,000 might become moot.

drofnad
11th January 2020, 09:31
I imagine that scaling might reduce the difference quite a bit...

I recall some remarks a long time ago from two users of, resp. Canon 1Ds & 5D(1) and Nikon D3X & D700, where each remarked that downsizing the larger-resolution (21/24mpx) files to match the (12mpx) smaller ones gained not only better noise handling but improved sharpness as well --the one fellow wondering if one was getting the full 12mpx of the D700. (I.e., that the downsized 21/24 mpx images were better than the 12mpx originals.)

YMMV?

-d.

ron787
11th January 2020, 10:14
I recall some remarks a long time ago from two users of, resp. Canon 1Ds & 5D(1) and Nikon D3X & D700, where each remarked that downsizing the larger-resolution (21/24mpx) files to match the (12mpx) smaller ones gained not only better noise handling but improved sharpness as well --the one fellow wondering if one was getting the full 12mpx of the D700. (I.e., that the downsized 21/24 mpx images were better than the 12mpx originals.)

YMMV?

-d.
Interesting note, but the unscientific test that I'd described had not involved any downsizing, just a small amount of noise reduction in C1. I do not have an S1, only an S1r, but it would be useful if someone with both the S1 and S1r could reproduce what I'd done with noise reduction under varying photographic situations to see if my findings hold up. Based upon the RAW samples provided by the link, I am leaning toward the impression that up to ISO 25,000, any extra noise encountered in an S1r image can be compensated for with noise reduction in C1, or perhaps other PP software, unless I can be convinced otherwise.

SrMphoto
12th January 2020, 07:33
Thank you for the RAW file link.
I've downloaded the comparative raw files from between ISO 6400 to 25,000 and opened them in Capture One Pro 20. The most noticeable differenceóto my eyes and with my calibrated monitoróappears between 12,000 (where 12,000 might be usable from both cameras. s1 more so than S1r) and 25,000, with the least noise apparent in the S1 samples. However, while there is a difference, it is not as significant as I'd anticipated, and I wonderófor myselfóhow often I'd be required to use ISO 12,000 or for that matter, 25,000.
When faced with a photographic opportunity that knowingly requires added light, I would bring along a flash and a fast lens. But, admittedly, there are venues where such artificial light sources might be distracting and or disallowed, and in that case C!'s noise reduction feature might come in handy for the S1r file.
I'm still on the fence in re the addition of an S1 to my S1r, I simply have to determine whether the very high ISO differential is worth the loss of detail afforded by the large sensor.
Thanks again.
Ron

P.S. I've just completed a comparison of the S1raw ISO 25000 file viewed at 100% in C1against the S1r raw file at 100% with C1's noise reduction set to #30 and aside from the fact that the S1r's image is larger, they reveal the same degree of limited noise as well as the same degree of sharpness (noise reduction tends to reduce sharpness to some degree depending upon the intensity employed). There is no visible color alteration. If this approach were to hold true for images acquired under other circumstances the S1 vs S1r High ISO differential up to ISO 25,000 might become moot.

Thanks for the report. Did you scale the images to the same dimensions? What I wonder is whether one can apply more noise reduction with higher resolution images before it damages the picture? That would explain why S1R's high ISO JPGs look better than S1's JPG images (i.e., more in-camera noise reduction applied with S1R than with S1),

ron787
12th January 2020, 08:26
Thanks for the report. Did you scale the images to the same dimensions? What I wonder is whether one can apply more noise reduction with higher resolution images before it damages the picture? That would explain why S1R's high ISO JPGs look better than S1's JPG images (i.e., more in-camera noise reduction applied with S1R than with S1),

Hi,
I did not scale the images, just downloaded the S1r and S1 RAW files as they were and imported them into C1 Pro 20 as is.
I do not have an answer, in re in camera noise reduction, although your assessment may be correct, but I was able to equalize the noise characteristics of the S1r's files to those of the S1 at high ISO by the judicious use of C1's noise reduction slider.

As an aside, I've done some testing with my GFX100, and it turns out to be a low light delight, enabling me to pull an image out of near blackness with little or no noise, but that is a completely different animal and, admittedly, entirely off topic.

ron787
13th January 2020, 12:07
I've done some further, more in depth testing of the S1r at high ISO, and up to ISO 3200 it is fine, and acceptable at 6400.
The noise at 6400 can be mitigated to some extent in C1 with noise reduction, and the degree of effectiveness appears to vary with the imaged subject. The end result being that the S1r, IMO, does not have a reliable, very low light imaging sensor, and I suspect that it had not been intended to be so. It is still one of my favorite all around cameras, but for those situations that call for very low light image acquisition, as well as occasional video, I have ordered a Panasonic S1H.

I understand that it is a video-centric camera and is overkill for a primarily stills photographer, but it has all of the S1/r's photo features, as well as the dual native ISO, which was, for myself, the deciding factor when considering between the S1vs S1H. And while it does incorporate an AA filter, the with AA and without examples that I've seen are hard to differentiate. So, we'll see how that goes and I'll post my thoughts after I've had a chance to give it a go around.

iiiNelson
14th January 2020, 02:45
I've done some further, more in depth testing of the S1r at high ISO, and up to ISO 3200 it is fine, and acceptable at 6400.
The noise at 6400 can be mitigated to some extent in C1 with noise reduction, and the degree of effectiveness appears to vary with the imaged subject. The end result being that the S1r, IMO, does not have a reliable, very low light imaging sensor, and I suspect that it had not been intended to be so. It is still one of my favorite all around cameras, but for those situations that call for very low light image acquisition, as well as occasional video, I have ordered a Panasonic S1H.

I understand that it is a video-centric camera and is overkill for a primarily stills photographer, but it has all of the S1/r's photo features, as well as the dual native ISO, which was, for myself, the deciding factor when considering between the S1vs S1H. And while it does incorporate an AA filter, the with AA and without examples that I've seen are hard to differentiate. So, we'll see how that goes and I'll post my thoughts after I've had a chance to give it a go around.

Apparently the S1 has dual native circuitry as well. Itís just not advertised.

https://fstoppers.com/gear/does-panasonic-s1-secretly-have-dual-native-iso-just-s1h-411220

ron787
14th January 2020, 05:19
Apparently the S1 has dual native circuitry as well. Itís just not advertised.

https://fstoppers.com/gear/does-panasonic-s1-secretly-have-dual-native-iso-just-s1h-411220

I had viewed that video and found this as well, "Trace is not completely sure of his findings as yet and needs to do some more tests. He notes that this only seems to occur when using the V-Log picture profile".

It appears that the S1 requires the V-log software key for the dualóassuming that his findings are to be verifiedóto be operational and thus it is not clear if the dual native ISO is available while shooting stills, while the S1H's dual native ISO is allegedly active at all times. But the dual native ISO is not the only advantageous feature of the S1H. That said, while there are many S1 owner/users out in the wild, I find it odd that no one else has made mention of the dual ISO, unless those individuals have avoided purchasing the V-log key.

iiiNelson
14th January 2020, 05:33
I had viewed that video and found this as well, "Trace is not completely sure of his findings as yet and needs to do some more tests. He notes that this only seems to occur when using the V-Log picture profile".

It appears that the S1 requires the V-log software key for the dualóassuming that his findings are to be verifiedóto be operational and thus it is not clear if the dual native ISO is available while shooting stills, while the S1H's dual native ISO is allegedly active at all times. But the dual native ISO is not the only advantageous feature of the S1H. That said, while there are many S1 owner/users out in the wild, I find it odd that no one else has made mention of the dual ISO, unless those individuals have avoided purchasing the V-log key.

Yeah I also believe that there are some tangible advantages to the S1H (which is why itís on my radar as well). Just wanted to pop that into the conversation because I didnít know if youíd seen those rumors.

ron787
14th January 2020, 12:58
Yeah I also believe that there are some tangible advantages to the S1H (which is why it’s on my radar as well). Just wanted to pop that into the conversation because I didn’t know if you’d seen those rumors.

The S1H arrived a few hours ago and I've only had a short time to set it up and fire off a few snaps, but it is virtually clean at ISO 6400 ... even 20,000. Pretty amazing. However, I'd acquired a few images from my home (behind Low E—metallic coated— impact glass) to the opposite side of the golf course upon which it is situated, and the loss of detail, as compared to the S1r when acquiring the same image with the same lens, is considerable. Of course, that's the 24MP sensor showing its limitations. But I had not ordered the S1h for landscape photography, and the comparative had been borne out of curiosity. That said, I've tried to add an image showing the absence of noise at ISO 6400 that had been cropped 100% but could not figure out how to accomplish the task on this site, sorry.

pegelli
14th January 2020, 23:59
:worthless:

ron787
15th January 2020, 04:56
:worthless:

Nice photos.
It appears that the only way to post to this site is via a URL, is that correct?

pegelli
15th January 2020, 05:59
Nice photos.
It appears that the only way to post to this site is via a URL, is that correct?

RTFM ;) : https://www.getdpi.com/forum/about-getdpi-rules-and-registration-information-read-this-if-you-are-new-/54-how-post-images-threads-use-gallery.html

ron787
15th January 2020, 09:00
RTFM ;) : https://www.getdpi.com/forum/about-getdpi-rules-and-registration-information-read-this-if-you-are-new-/54-how-post-images-threads-use-gallery.html

Thank you for the link.