Application of Reflectance Transformation Imaging to Experimental Archaeology Studies

Round 1

Reviewer 1 Report

In this study, the techinques of RTI and RSI are integrated for the in-situ 3D reconstructions of archaeological items with micrometric resolution. The work presented is interesting for the field archaeology. My personnel comments for the further improvement of this manuscript can be summarized into twofold:

1) the techinical issues w.r.t the applied technologies of RTI and RSI need to be clarified in more detail. Firstly, why the RTI technique is benefical for the resolution enhancement? What is its performance compared with conventional photogrammetry and digital microscope, in particular using the assessment index of spatial resolution. Second, how RTI and RSI are integated, and how the depth profile with micrometric accuracy can be derived.
2) the data processing time for the 3D reconstruction of the archaeological stick need to be added in the revision in order to present the feasibility of the applied technology and algorithm. Of course, the applied parameters for the data processing is also recommended to be provided.

Author Response

In this study, the techinques of RTI and RSI are integrated for the in-situ 3D reconstructions of archaeological items with micrometric resolution. The work presented is interesting for the field archaeology. My personnel comments for the further improvement of this manuscript can be summarized into twofold:

• the techinical issues w.r.t the applied technologies of RTI and RSI need to be clarified in more detail. Firstly, why the RTI technique is benefical for the resolution enhancement? What is its performance compared with conventional photogrammetry and digital microscope, in particular using the assessment index of spatial resolution. Second, how RTI and RSI are integated, and how the depth profile with micrometric accuracy can be derived.

In the revised version of the manuscript we clarified the point that RTI provides a resolution comparable with the one of digital microscopy. However, the possibility of varying interactively the illumination angle makes RTI advantageous with respect to digital microscopy. For the metric analysis of the depth profiles, which is done exploiting the RTI images and the RSI algorithm, the comparison is with conventional photogrammetry, which hardly has resolutions better than 10 microns. The RSI method allows a metric reconstruction; knowing the linear dimensions of the object under study, all the other dimensions can be scaled accordingly in the digital model.   

2) the data processing time for the 3D reconstruction of the archaeological stick need to be added in the revision in order to present the feasibility of the applied technology and algorithm. Of course, the applied parameters for the data processing is also recommended to be provided.

The RTI image is acquired and generated in less than five minutes. The elaboration time for reconstructing the 3D profile of the surface depends on the area to be analysed, it can range from a few minutes for small areas (a few mm2) to over one hour for large surfaces (several cm2). In our case, the RSI reconstruction took about 30 minutes on a average-performance laptop computer (using 12 images, at maximum resolution, on a surface of about 5 cm2)

Reviewer 2 Report

The authors present a case of using Reflectance Transformation Imaging (RTI) for the documentation of use wear analysis of a replica wooden digging stick. The paper has some merits, but needs substantial revision/amplification.

The authors place their work in a research vacuum. From reading the paper it appears that thgeir ork is the first of its kid. For from it. There is a large body of literature on the use of RTI in archaeological, monumental and curatorial contexts with 740 references since 2016 alone

https://0-scholar-google-com.brum.beds.ac.uk/scholar?hl=en&as_sdt=0%2C5&as_ylo=2016&q=%22Reflectance+Transformation+Imaging%22+archaeology&btnG=

The authors reference only a single paper. There needs to be a better review of the general RTI literature than provided. The literature review should then covers wooden objects or other organics.

Particularly relevant papers are, inter alia:

Rich, Sara A., Ryan Watts, and Garry Momber. "Mesolithic woodworking, experimental archaeology & underwater heritage in Hampshire and the Isle of Wight (UK)." Mesolithic Miscellany 24, no. 1 (2016): 3-12.

Newman, Sarah E. "Applications of reflectance transformation imaging (RTI) to the study of bone surface modifications." Journal of Archaeological Science 53 (2015): 536-549.

I wonder whether the MATLAB script should be reproduced as a table in the paper?

The discussion is very perfunctory and cursory…it needs to tie back to the known literature and show the advances made by the authors.

On line 161 it is claimed to be “quite complex and time consuming”, while on line 169 the authors claim suitability for “quick in situ documentation and assessment” Which is it?

Line 161-162  How much time is needed? If fast enough, the RTI method seems to offer an opportunity to carry out an experiment, look at the resulting use-wear, and then modify the experimental approach if the technique can be adjusted to become near real time, it would be a powerful tool.

On a formal level, this paper needs a thorough edit by a native English speaker.

The papers cited in reference list need to be checked for proper capitalisation of the titles

Author Response

The authors present a case of using Reflectance Transformation Imaging (RTI) for the documentation of use wear analysis of a replica wooden digging stick. The paper has some merits, but needs substantial revision/amplification.

The authors place their work in a research vacuum. From reading the paper it appears that thgeir ork is the first of its kid. For from it. There is a large body of literature on the use of RTI in archaeological, monumental and curatorial contexts with 740 references since 2016 alone

https://0-scholar-google-com.brum.beds.ac.uk/scholar?hl=en&as_sdt=0%2C5&as_ylo=2016&q=%22Reflectance+Transformation+Imaging%22+archaeology&btnG=

The authors reference only a single paper. There needs to be a better review of the general RTI literature than provided. The literature review should then covers wooden objects or other organics.

Particularly relevant papers are, inter alia:

Rich, Sara A., Ryan Watts, and Garry Momber. "Mesolithic woodworking, experimental archaeology & underwater heritage in Hampshire and the Isle of Wight (UK)." Mesolithic Miscellany 24, no. 1 (2016): 3-12.

Newman, Sarah E. "Applications of reflectance transformation imaging (RTI) to the study of bone surface modifications." Journal of Archaeological Science 53 (2015): 536-549.

The discussion on the state of the art has been improved, following the suggestions of the reviewer, and the suggested references were added.

I wonder whether the MATLAB script should be reproduced as a table in the paper?

The script we wrote implemented the Depth Gradient Fusion package developed by the University of Toulouse. Me added a link to the example Matlab scripts provided by the authors, the same that we used after adapting to our specific setup.

The discussion is very perfunctory and cursory…it needs to tie back to the known literature and show the advances made by the authors.

We have extended the discussion of the results, added some additional references to applications of the RTI technique to woodden archaeological objects, as per the reviewer suggestions

On line 161 it is claimed to be “quite complex and time consuming”, while on line 169 the authors claim suitability for “quick in situ documentation and assessment” Which is it? Line 161-162  How much time is needed? If fast enough, the RTI method seems to offer an opportunity to carry out an experiment, look at the resulting use-wear, and then modify the experimental approach if the technique can be adjusted to become near real time, it would be a powerful tool.

We have clarified that the RTI image can be acquired in a very short time The acquisition and elaboration of the RTI image is very fast, it takes less than 5 minutes with our system, which is not optimized for speed.. On the other hand, the metrical reconstruction of the surface profile with the RSI technique takes much time (especially when very high spatial resolution is required). The time needed also depends on the area to be analysed, it can range from a few minutes for small areas (a few cm2) to over one hour for large surfaces.

On a formal level, this paper needs a thorough edit by a native English speaker.

We have revised and improved the English where needed

The papers cited in reference list need to be checked for proper capitalisation of the titles

We have checked and revised the capitalization of the titles in the references

Round 2

Reviewer 1 Report

The comments raised in the 1st round review has been carefully considered in the revision, and thus it can be accepted in this form. 

Author Response

We would like to thank the reviewer for his/her suggestions

Reviewer 2 Report

The authors have addressed some issues raised by  my and by the other reviewer, and that is commendable.

However, the discussion of existing literature is still very inadequate and reads very cursorily. While the authors have added a couple of  references in the background section (line 62), and have substantially improved their section 2, the overall  contextualisation is still not clear. Without adequate context the significance of the research cannot be assessed.

As I noted in the initial review, the research undertaken by the authors does nit exist in a vacuum. I would have expected that the authors present a full context of the method as applied to wooden and other organic objects and to discuss the known opportunities and shortcomings of existing research.  This then grounds the work they are conducted and are reporting. All the authors have done is cite the two papers  that I highlighted in my initial review. The authors did not engage with the content of the papers. And, beyond that, there are additional papers to be considered.

The discussion is still very perfunctory and cursory…it needs to tie back to the known literature (which has not been presented) and show the advances made by the authors. At this point in time, even after the revision, I am wondering where the 'newsworthiness' and scientific advances of the paper rests. The authors have not made clear why this is worth publishing.

I my opinion, the authors are doing themselves an enormous disfavour by not addressing this properly. Can I encourage the authors to undertake another, poroper revision? It will strengthen the paper and make it citable for other researchers.

Author Response

We would like to thank the reviewer for his/her suggestions. Although our work is not intended to be a review on the applcations of the RTI technique, we agree that a more detailed discussion of the existing literature is appropriate. We moved all this part to the Introduction, while in Section 2 we discussed the experimental setup and the results. We hope that in this form our paper could be suitable for publication on Heritage.