The GetDPI Photography Forum

Great to see you here. Join our insightful photographic forum today and start tapping into a huge wealth of photographic knowledge. Completing our simple registration process will allow you to gain access to exclusive content, add your own topics and posts, share your work and connect with other members through your own private inbox! And don’t forget to say hi!

Early X100 Observations

barjohn

New member
Of course no camera is perfect and the X100 is no exception. Before I get into the positives, I'll dispense with the negatives, most of which can be fixed in firmware.

1. We need to have the ability to have another programmable button. The RAW button is a natural candidate as it is a function that could as easily serve in a menu for frequency of use. This button would be FN2 and most likely be used for turning on or off the ND filter. However, the photographer could choose as he does for FN1.
2. Auto ISO should be moved to the ISO menu and its parameters selected there with out limitation.
3. Multi-thread the control functions such that they can continue to be used while an image is being written to the memory card.
. Dramatically speed up the write of RAW files to SD memory. This function is way, way too slow and given the other performance parameters cannot be a hardware limitation but a programming one where the writes of RAW are processed in some arcane way that takes too long. With a fast processor and a high speed data path it shouldn't take so long.
4. When you select the focus rectangle size it goes back to the default after the shot. I want to pick one and have it stay until I decide to change it. Same is true for going into macro mode. I want it to stay in macro mode for multiple shots until I no longer want it in macro mode. Better yet, give me both. Option 1 only for next shot, option 2 stay until changed. Same for size of focus rectangle.

5. You cannot change the size of the focus rectangle in the OVF and you can't perform the function when the camera is up to your face.

These may appear to be minor quibbles but they would enhance usability greatly and are all simple firmware changes. As I get to know the camera there may be more.

Now to the good:

This camera lens holds its own and I think, based on both first hand observations and the tests performed by others out performs the X! which I considered a benchmark in lens performance for these small APSC cameras. Since I don't have an X1, I could only note for myself how sharp the images and how high the resolution the images are when I compared them to other cameras I have owned. But this link to a test by a professional photographer that really compares the two lenses in the real world of a photo shoot is the most convincing. Additionally, this comparison between the X100 and X1 and 5D II provides both useful image comparisons but more importantly the difference in distortion is clearly visible.

Finally this video comparing AF speeds between a GF2, X100 and an X1, while not purely scientific gives a good relative performance view.

The bottom line is that I believe this camera beats out the X1 at its strong suit, costs at least $800 less and offer features the X1 doesn't have and can't offer. I just hope that Fuji is able to deliver them with the chaos currently in Japan. They will need all of the business they can get to rebuild their economy and to try and recover from this disaster.
 
Last edited:

Terry

New member
John,
Would you please expand on number 3. I'm not sure what you mean by multi thread? At first I thought there is no impact to camera function during the write time. Then I heard that you can shoot more, what you don't have is detailed review. I'm not sure what your post is saying you have or don't have while the file is writing to the card?
 

glenerrolrd

Workshop Member
Am I missing something in these tests? The exposure values of the test images seem to be sufficiently off to distort any chance at a comparison. The Leica signature seems to show the color clarity and micro contrast that I am familiar with. The x100 images look muddy in some examples.

as we have seen with the S2 the processing profiling will have a huge impact on the utlimate rendering.. I will be very surprised if the X100 isn t able to be compatible to the X1 using best practices.

The real attraction is the form and handling. the viewfinder ,manual focusing etc fit the street shooting most requested list. AF will never match a Dslr for either speed or accuracy...nor can it compare to RF MF on a M body. The x1 is difficult to use for street work( see Sean Reids tests).

Of course price and the cool factor also make the X100 attractive...but if doesn t deliver benefits in form and handling you will see many onthe resale list this summer.
 

Terry

New member
AF will never match a Dslr for either speed or accuracy...nor can it compare to RF MF on a M body.
While Phase detect may be faster it isn't necessarily more accurate. You don't have to worry about the front focus/back focus issues of dslr lenses coupled with phase detect (think: lens micro adjustments that can be needed on dslrs).

So far from everything that I've read, yes there are some things that would be good for firmware upgrades but no deal breakers anywhere to say that it isn't a very good street shooter.
 

barjohn

New member
John,
Would you please expand on number 3. I'm not sure what you mean by multi thread? At first I thought there is no impact to camera function during the write time. Then I heard that you can shoot more, what you don't have is detailed review. I'm not sure what your post is saying you have or don't have while the file is writing to the card?
Terry,

Most cameras these days use a variant of the open source Lynux Operating System (OS). I don't know whether Fuji does or not, but even if they are using their own proprietary OS, virtually all modern OSs provide both multi-tasking and multi-threading capabilities. This is how they give you the perception that they are doing multiple things at the same time in that they divide time into discrete slices and allocate that time to multiple tasks rapidly switching from one to the next. There are many subtle variations on the model that I won't bore you with here. The task itself can be further split into threads that are event driven and spawned. So while it is performing the write from buffer to SD memory task, the event of you pushing a button spawns a sub-task to perform the buttons function. If the sub-task does not require the same resources or those resources have been written so they are sharable then the sub-task is executed and the user sees the camera as being responsive. If the answer to the former is no to either question, then the user gets no response to her action.

Most modern applications give priority to user actions over background tasks and that is what makes the difference between a camera that the user perceives to be responsive and one that the user thinks is unresponsive.

Using a stop watch I timed Fuji's write times with a class 10 card. The results are as follows:

Fine Only 2.5 sec
RAW Only 3.5 Sec
RAW + Fine 6 Sec

We know that the X100 is capable of multi-tasking because you can continue to shoot pictures (going to the buffer memory) while it is writing from the buffer memory to the SD card. This also tells us that the buffer can be read from and written to concurrently (most modern memories support this capability). It also tells us that the image processor can concurrently process a new image creating a RAW and JPG while the older image is being written to the SD card. There are different ways to move data from one memory to another and different capabilities for the pipeline. Since one RAW + JPG is about 128 mbits and it takes 6 seconds to write that is a rate of 21.33 mbits per sec or 2.67 mbytes per second. We know the SD card is capable of much higher write speeds so there is a problem in how the data is being written. Fuji should be able to improve this performance substantially. My guess is that the way they are doing the writes is using the slowest method and checking for errors with each write rather than performing a block write and checking for errors at the end. However, even if they chose not to improve write speed, they could allow all user interactions that involve using buffer memory to proceed by sharing the reads to buffer memory (this would slow the write down a little but the user wouldn't know or care) and/or allow other functions that don't require reads to buffer memory to proceed.

This is probably more than you wanted to know but I hope it helps.

Finally, I think I have found a bug in the AF software. I was playing with AF and holding the camera as still as possible and repeatedly pushing the shutter release 1/2 way to obtain focus confirmation. I found that the distance displayed would change on a random basis by several feet. I need to further test this with a tripod to make sure that it wasn't due to my moving the focus point. I will report my findings later.
 

Terry

New member
I understand what you are writing. Other than the camera finishing it's write time faster what will change in camera useability?
 

Bob

Administrator
Staff member
This is a refreshing potentially visionary camera.
I can't wait to play with one.
-bob
 

barjohn

New member
I understand what you are writing. Other than the camera finishing it's write time faster what will change in camera useability?
You could review an image as soon as it was processed (much faster than the write time to SD.). You could change settings, currently all of the various controls are locked during the write to SD card. Therefore if you want to change shutter speed, WB, to Macro, AF points, etc. you can't until writing is complete. This makes the camera seem less responsive as you can't prepare for the next shot until all of the prior images are written. Interestingly, if auto review is set to on, such as the 1.5 sec review, it appears instantly after taking the shot while the image is being written. However, you can't zoom the image. There seems to be an inconsistency in the logic.

My guess, they were in a hurry to get this out so they took a few shortcuts with plans to address these shortcoming in a future release as they see how users respond.
 

Terry

New member
Thanks,
But then begs more questions,
If the aperture ring and shutter are mechanical dials what happens if you change them when the card is writing?
If you point the camera at the sky and take a shot and then immediaty point the camera down to the ground and take a shot, is the exposure on the second shot correct?
 

barjohn

New member
Nothing happens until the write is complete and now you can't take another shot either. Once complete it changes to the new setting. With no changes, you can shoot the sky then immediately shoot the ground without changing any setting and get the right exposure for each shot assuming that either shutter speed or aperture is set to A or both are set to A.
 
Nothing happens until the write is complete and now you can't take another shot either. Once complete it changes to the new setting. With no changes, you can shoot the sky then immediately shoot the ground without changing any setting and get the right exposure for each shot assuming that either shutter speed or aperture is set to A or both are set to A.
This is not what I am seeing. I just went outside and shot five consecutive frames, quickly changing the aperture between each one while the card was writing. There was no delay in shooting and the aperture changed as would be expected.

Maybe I'm not clear on what barjohn is referring to.
 
Ahh, I think I see. He was referring to changing some menu driven settings and is correct but I think he confused Terry's point since you can in fact change the mechanical settings while the card is writing.
 
And I agree with the assessment that the firmware was likely rushed out to meet the ship date. I fully expect that we are going to see some meaningful improvements to the "oddities" in the current firmware.
 
Thanks,
But then begs more questions,
If the aperture ring and shutter are mechanical dials what happens if you change them when the card is writing?
If you point the camera at the sky and take a shot and then immediaty point the camera down to the ground and take a shot, is the exposure on the second shot correct?
Terry always asks the right questions... :)
 

Terry

New member
Ahh, I think I see. He was referring to changing some menu driven settings and is correct but I think he confused Terry's point since you can in fact change the mechanical settings while the card is writing.
In normal shooting I don't seem to need to get to the menus very often as I use aperture or shutter priority most of the time. I'm also generally ok with auto ISO except when doing landscapes where I lock down the iso to the base value.

I'm just trying to figure out what are firmware improvements that will be helpful vs show stoppers in the current state. This to me seems like it would be a good thing to workout in a firmware update but doesn't render the cera a dud in it's current state.
 

barjohn

New member
I stand corrected. When I tried it the first time, it did not let me change aperture and shoot. I just tried it again and it did. I'm not sure what I am doing different. I don't think any of the things I have mentioned are in the "show stopper" category but rather in the enhance usability.
 
In normal shooting I don't seem to need to get to the menus very often as I use aperture or shutter priority most of the time. I'm also generally ok with auto ISO except when doing landscapes where I lock down the iso to the base value.

I'm just trying to figure out what are firmware improvements that will be helpful vs show stoppers in the current state. This to me seems like it would be a good thing to workout in a firmware update but doesn't render the cera a dud in it's current state.
Here you go Terry...

http://www.x100forum.com/index.php?/topic/154-laundry-list-of-general-issues-with-ver-1-firmware/
 
Top